ISP Meets Deep Learning: A Survey on Deep Learning Methods for Image Signal Processing
Open accessAbstract
The entire Image Signal Processor (ISP) of a camera relies on several processes to transform the data from the Color Filter Array (CFA) sensor, such as demosaicing, denoising, and enhancement. These processes can be executed either by some hardware or via software. In recent years, Deep Learning (DL) has emerged as one solution for some of them or even to replace the entire ISP using a single neural network for the task. In this work, we investigated several recent pieces of research in this area and provide deeper analysis and comparison among them, including results and possible points of improvement for future researchers.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
Cites
U-Net: Convolutional Networks for Biomedical Image Segmentation
Ronneberger, O., Fischer, P., & Brox, T. (2015) · Lecture notes in computer science · doi:10.1007/978-3-319-24574-4_28 · Cited by 99,404 · cited 6× · see citing works
“In the first, results were validated on the Samsung S7 dataset [70], and CURL scored the best PSNR and LPIPS metrics when compared to the U-Net [85] and DeepISP [70] methods, but tied with the DeepISP method on the SSIM metric.”
InfluentialDensely Connected Convolutional Networks
Huang, G., Liu, Z., Van Der Maaten, L., & Weinberger, K. Q. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.243 · Cited by 48,566 · cited 1× · see citing works
“It also solved the vanishing-gradient problem experienced by many deep neural networks, due to the residual learning [73] and densely connected convolutional neural network [74].”
ImageNet Large Scale Visual Recognition Challenge
Russakovsky, O., Deng, J., Su, H., Krause, J., Satheesh, S., Ma, S., Huang, Z., Karpathy, A., Khosla, A., Bernstein, M., et al. (2015) · International Journal of Computer Vision · doi:10.1007/s11263-015-0816-y · Cited by 39,380 · cited 2× · see citing works
“The first was by discarding denoising and de-mosaicing pre-processing on a model pre-trained on the ImageNet dataset [96].”
MobileNetV2: Inverted Residuals and Linear Bottlenecks
Sandler, M., Howard, A., Zhu, M., Zhmoginov, A., & Chen, L.-C. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00474 · Cited by 28,755 · cited 2× · see citing works
“ResNet18 [56] and MobileNetV2 [97] were used for validation.”
Image-to-Image Translation with Conditional Adversarial Networks
Isola, P., Zhu, J.-Y., Zhou, T., & Efros, A. A. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.632 · Cited by 22,930 · cited 2× · see citing works
“The first compared PyNET with the methods SPADE [85], DPED [45], U-Net [93], Pix2Pix [51], SRGAN [66], VDSR [63], and SRCNN [23].”
A survey on deep learning in medical image analysis
Litjens, G., Kooi, T., Bejnordi, B. E., Setio, A. A. A., Ciompi, F., Ghafoorian, M., van der Laak, J. A. W. M., van Ginneken, B., & Sánchez, C. I. (2017) · Medical Image Analysis · doi:10.1016/j.media.2017.07.005 · Cited by 15,332 · cited 1× · see citing works
“In the field of Deep Learning, for example, a great number of surveys is already available in many different areas, such as agriculture [15], cyber security [16], autonomous driving [17], medical imaging [18], and also on more technical areas, such as on CNN [19].”
Are we ready for autonomous driving? The KITTI vision benchmark suite
Geiger, A., Lenz, P., & Urtasun, R. (2012) · 2012 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2012.6248074 · Cited by 14,897 · cited 1× · see citing works
“VisionISP was trained and evaluated with the KITTI 2D object detection dataset [79], an autonomous driving benchmark dataset.”
The Unreasonable Effectiveness of Deep Features as a Perceptual Metric
Zhang, R., Isola, P., Efros, A. A., Shechtman, E., & Wang, O. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00068 · Cited by 14,623 · cited 2× · see citing works
“Moreover, the GAN version of this model obtains better perceptual results in the LPIPS metric[107].”
Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network
Ledig, C., Theis, L., Huszar, F., Caballero, J., Cunningham, A., Acosta, A., Aitken, A., Tejani, A., Totz, J., Wang, Z., et al. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.19 · Cited by 12,696 · cited 4× · see citing works
“The first compared PyNET with the methods SPADE [85], DPED [45], U-Net [93], Pix2Pix [51], SRGAN [66], VDSR [63], and SRCNN [23].”
InfluentialDecoupled Weight Decay Regularization
Loshchilov, I., & Hutter, F. (2017) · arXiv · doi:10.48550/arxiv.1711.05101 · Cited by 12,522 · see citing works
The perceptron: A probabilistic model for information storage and organization in the brain.
Rosenblatt, F. (1958) · Psychological Review · doi:10.1037/h0042519 · Cited by 12,280 · cited 4× · see citing works
“Although modern image sensors are capable to fix many data acquisition issues[4], a pre-processing of the data contained in the RAW image is important to verify signal integrity, identify and fix acquisition failures, preventing errors from being propagated through the following operations.”
InfluentialPerceptual Losses for Real-Time Style Transfer and Super-Resolution
Johnson, J., Alahi, A., & Fei-Fei, L. (2016) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-46475-6_43 · Cited by 11,283 · cited 2× · see citing works
“Intermediate levels are trained by a combination of MSE and VGG-based [54] to refine the color and shape of objects.”
Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising
Zhang, K., Zuo, W., Chen, Y., Meng, D., & Zhang, L. (2017) · IEEE Transactions on Image Processing · doi:10.1109/tip.2017.2662206 · Cited by 8,983 · cited 2× · see citing works
“For example, deep learning models can achieve state-of-the-art performance in tasks like denoising [87], demosaicing [46], and super-resolution [133], often surpassing classical algorithms.”
Martin, D., Fowlkes, C., Tal, D., & Malik, J. (2002) · Proceedings Eighth IEEE International Conference on Computer Vision. ICCV 2001 · doi:10.1109/iccv.2001.937655 · Cited by 8,216 · cited 4× · see citing works
“The results were validated on the MSR demosaicing dataset [125], BSD100 [113], McMaster [83], Urban 100 [84], Kodak [82], and WED [126] datasets and compared with Deepjoint4 [60], Kokkinos [96], Dong [127], DeepISP [70], and DPN [29] methods at three different noise levels (5, 15, 25).”
InfluentialAccurate Image Super-Resolution Using Very Deep Convolutional Networks
Kim, J., Lee, J. K., & Lee, K. M. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.182 · Cited by 7,870 · cited 2× · see citing works
“The first compared PyNET with the methods SPADE [85], DPED [45], U-Net [93], Pix2Pix [51], SRGAN [66], VDSR [63], and SRCNN [23].”
Deformable Convolutional Networks
Dai, J., Qi, H., Xiong, Y., Li, Y., Zhang, G., Hu, H., & Wei, Y. (2017) · 2017 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2017.89 · Cited by 7,513 · cited 1× · see citing works
“The InterF uses a deformable convolution [121] in the feature domain to make up for the shift between frames.”
Grad-CAM: Visual Explanations from Deep Networks via Gradient-Based Localization
Selvaraju, R. R., Cogswell, M., Das, A., Vedantam, R., Parikh, D., & Batra, D. (2020) · International Journal of Computer Vision · doi:10.1007/s11263-019-01228-7 · Cited by 6,909 · cited 1× · see citing works
“Even though there are methods for interpretability in Deep Learning [103, 116], these methods might not help specifically in the end-to-end ISP pipeline based on Deep Learning methods because they would not highlight the areas that contribute with errors.”
Multiscale structural similarity for image quality assessment
Wang, Z., Simoncelli, E. P., & Bovik, A. C. (2004) · The Thrity-Seventh Asilomar Conference on Signals, Systems & Computers, 2003 · doi:10.1109/acssc.2003.1292216 · Cited by 6,253 · cited 1× · see citing works
“Finally, level one is trained with MSE, VGG, and SSIM loss [88] and performs corrections in the local color processing, noise removal, texture enhancement, etc.”
Image Super-Resolution Using Very Deep Residual Channel Attention Networks
Zhang, Y., Li, K., Li, K., Wang, L., Zhong, B., & Fu, Y. (2018) · Lecture notes in computer science · doi:10.1007/978-3-030-01234-2_18 · Cited by 6,003 · cited 3× · see citing works
“An important aspect of this work is its loss function, which combines Charbonnier loss [138], SSIMloss, and Perceptual loss tominimize perceptual differences between the generated image and the ground truth.”
Deep learning in agriculture: A survey
Kamilaris, A., & Prenafeta-Boldú, F. X. (2018) · Computers and Electronics in Agriculture · doi:10.1016/j.compag.2018.02.016 · Cited by 5,579 · cited 1× · see citing works
“In the field of Deep Learning, for example, a great number of surveys is already available in many different areas, such as agriculture [15], cyber security [16], autonomous driving [17], medical imaging [18], and also on more technical areas, such as on CNN [19].”
ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks
Wang, X., Yu, K., Wu, S., Gu, J., Liu, Y., Dong, C., Qiao, Y., & Loy, C. C. (2019) · Lecture notes in computer science · doi:10.1007/978-3-030-11021-5_5 · Cited by 5,085 · cited 3× · see citing works
“With the SR-RAW dataset, the authors also compared with SR methods, as SRGAN[92], ESRGAN [108], SPSR [109], and RealSR[110].”
A Review of Image Denoising Algorithms, with a New One
Buades, A., Coll, B., & Morel, J. (2005) · Multiscale Modeling and Simulation · doi:10.1137/040616024 · Cited by 4,120 · cited 1× · see citing works
“Image denoising is very important for several applications in the vision computing field and has received a lot of attention over the years [53, 54, 55, 21, 56].”
Single image super-resolution from transformed self-exemplars
Huang, J., Singh, A., & Ahuja, N. (2015) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2015.7299156 · Cited by 3,603 · cited 7× · see citing works
“[84], in order to address the lack of HighResolution Datasets with indoor, urban, and architectural scenes.”
InfluentialImproved protein structure prediction using potentials from deep learning
Senior, A. W., Evans, R., Jumper, J., Kirkpatrick, J., Sifre, L., Green, T., Qin, C., Žídek, A., Nelson, A. W. R., Bridgland, A., et al. (2020) · Nature · doi:10.1038/s41586-019-1923-7 · Cited by 3,378 · cited 1× · see citing works
“For that reason, several areas such as Chemistry [104], Medicine [22], Economics [106], and Physics [5], could harness its capacities to accelerate or improve their work, directly impacted by this evolution in the field known as Deep Learning.”
Semantic Image Synthesis With Spatially-Adaptive Normalization
Park, T., Liu, M.-Y., Wang, T.-C., & Zhu, J.-Y. (2019) · 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2019.00244 · Cited by 2,974 · cited 2× · see citing works
“The first compared PyNET with the methods SPADE [85], DPED [45], U-Net [93], Pix2Pix [51], SRGAN [66], VDSR [63], and SRCNN [23].”
International Journal of Computer Vision
(2013) · International Journal of Computer Vision · doi:10.1007/11263.1573-1405 · Cited by 2,338 · see citing works
A survey of deep learning techniques for autonomous driving
Grigorescu, S., Trasnea, B., Cocias, T., & Macesanu, G. (2020) · Journal of Field Robotics · doi:10.1002/rob.21918 · Cited by 1,809 · cited 1× · see citing works
“In the field of Deep Learning, a great number of surveys is already available in many different areas, such as agriculture [58], biometrics [27], cyber security [7], regularization [98], autonomous driving [33], medical imaging [71], and also on more technical areas, such as on CNN [69].”
Score-CAM: Score-Weighted Visual Explanations for Convolutional Neural Networks
Wang, H., Wang, Z., Du, M., Yang, F., Zhang, Z., Ding, S., Mardziel, P., & Hu, X. (2020) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw50498.2020.00020 · Cited by 1,553 · cited 1× · see citing works
“Even though there are methods for interpretability in Deep Learning [103, 116], these methods might not help specifically in the end-to-end ISP pipeline based on Deep Learning methods because they would not highlight the areas that contribute with errors.”
Video Enhancement with Task-Oriented Flow
Xue, T., Chen, B., Wu, J., Wei, D., & Freeman, W. T. (2019) · International Journal of Computer Vision · doi:10.1007/s11263-018-01144-2 · Cited by 1,401 · see citing works
Chen, C., Chen, Q., Jia, X., & Koltun, V. (2018) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2018.00347 · Cited by 1,388 · see citing works
Deep Learning on Image Denoising: An overview
Tian, C., Fei, L., Zheng, W., Xu, Y., Zuo, W., & Lin, C.-W. (2020) · Neural Networks · doi:10.1016/j.neunet.2020.07.025 · Cited by 1,091 · cited 1× · see citing works
“Image denoising is very important for several applications in the vision computing field and has received a lot of attention over the years [10, 25, 81, 95, 110].”
Learning Enriched Features for Real Image Restoration and Enhancement
Zamir, S. W., Arora, A., Khan, S., Hayat, M., Khan, F. S., Yang, M., & Shao, L. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-58595-2_30 · Cited by 1,069 · cited 3× · see citing works
“This network, defined as a pixel-bin image processing network (PIPNet) , uses UNet as a framework and traverses different feature ISP Meets DL: A Survey on Deep Learning Methods for Image Signal Processing 127:9 depths through downscaling and upscaling operations.”
Underexposed Photo Enhancement Using Deep Illumination Estimation
Wang, R., Zhang, Q., Fu, C., Shen, X., Zheng, W., & Jia, J. (2019) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2019.00701 · Cited by 1,044 · cited 2× · see citing works
“In the second, results were validated over the MIT-Adobe FiveK dataset [11] and compared with HDRNet [31], DPE [16], White-Box [41], Distort-and-Recover [84], and DeepUPE [117] methods where CURL scored better on PSNR and LPIPS metrics, but DeepUPE scored the best on SSIM.”
A High-Quality Denoising Dataset for Smartphone Cameras
Abdelhamed, A., Lin, S. C.-F., & Brown, M. S. (2018) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2018.00182 · Cited by 911 · see citing works
Deep bilateral learning for real-time image enhancement
Gharbi, M., Chen, J., Barron, J. T., Hasinoff, S. W., & Durand, F. (2017) · ACM Transactions on Graphics · doi:10.1145/3072959.3073592 · Cited by 805 · cited 1× · see citing works
“In the second, results were validated over the MIT-Adobe FiveK dataset [11] and compared with HDRNet [31], DPE [16], White-Box [41], Distort-and-Recover [84], and DeepUPE [117] methods where CURL scored better on PSNR and LPIPS metrics, but DeepUPE scored the best on SSIM.”
Waterloo Exploration Database: New Challenges for Image Quality Assessment Models
Ma, K., Duanmu, Z., Wu, Q., Wang, Z., Yong, H., Li, H., & Zhang, L. (2017) · IEEE Transactions on Image Processing · doi:10.1109/tip.2016.2631888 · Cited by 789 · cited 2× · see citing works
“The results were validated on the MSR demosaicing dataset [125], BSD100 [113], McMaster [83], Urban 100 [84], Kodak [82], and WED [126] datasets and compared with Deepjoint4 [60], Kokkinos [96], Dong [127], DeepISP [70], and DPN [29] methods at three different noise levels (5, 15, 25).”
Real Image Denoising With Feature Attention
Anwar, S., & Barnes, N. (2019) · 2019 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2019.00325 · Cited by 784 · cited 3× · see citing works
“CycleISP achieved better results in both scenarios, with almost a 5 times smaller network parameters number than the previous best RAW denoising method [58] and performance gain against the previous best sRGB denoising algorithm [57].”
Brief review of image denoising techniques
Fan, L., Zhang, F., Fan, H., & Zhang, C. (2019) · Visual Computing for Industry Biomedicine and Art · doi:10.1186/s42492-019-0016-7 · Cited by 699 · see citing works
DSLR-Quality Photos on Mobile Devices with Deep Convolutional Networks
Ignatov, A., Kobyshev, N., Timofte, R., & Vanhoey, K. (2017) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2017.355 · Cited by 608 · cited 2× · see citing works
“The first compared PyNET with the methods SPADE [85], DPED [45], U-Net [93], Pix2Pix [51], SRGAN [66], VDSR [63], and SRCNN [23].”
Deep high dynamic range imaging of dynamic scenes
Kalantari, N. K., & Ramamoorthi, R. (2017) · ACM Transactions on Graphics · doi:10.1145/3072959.3073609 · Cited by 599 · cited 1× · see citing works
“To train the network, synthetic and real datasets based on Kalantari et al. [57] datasets were used.”
A Survey of Deep Learning Methods for Cyber Security
Berman, D. S., Buczak, A. L., Chavis, J. S., & Corbett, C. L. (2019) · Information · doi:10.3390/info10040122 · Cited by 569 · cited 1× · see citing works
“In the field of Deep Learning, a great number of surveys is already available in many different areas, such as agriculture [58], biometrics [27], cyber security [7], regularization [98], autonomous driving [33], medical imaging [71], and also on more technical areas, such as on CNN [69].”
Color demosaicking by local directional interpolation and nonlocal adaptive thresholding
Lei Zhang, A., Wu, X., Buades, A., & Li, X. (2011) · Journal of Electronic Imaging · doi:10.1117/1.3600632 · Cited by 552 · cited 13× · see citing works
“[83] and consists of eighteen sub-images of size 500x500 captured by Kodak film and then digitized.”
InfluentialLearning photographic global tonal adjustment with a database of input / output image pairs
Bychkovsky, V., Paris, S., Chan, E., & Durand, F. (2011) · doi:10.1109/cvpr.2011.5995332 · Cited by 531 · see citing works
Unprocessing Images for Learned Raw Denoising
Brooks, T., Mildenhall, B., Xue, T., Chen, J., Sharlet, D., & Barron, J. T. (2019) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2019.01129 · Cited by 493 · cited 6× · see citing works
“Besides, from these high-resolution RAW images was created the low-resolution RAW images through the bicubic downsampling kernel [94], mosaic kernel [58], and addition of the Gaussian-Poisson noise model [111].”
InfluentialBurst photography for high dynamic range and low-light imaging on mobile cameras
Hasinoff, S. W., Sharlet, D., Geiss, R., et al. (2016) · ACM Transactions on Graphics · doi:10.1145/2980179.2980254 · Cited by 483 · cited 3× · see citing works
“…PatchNet [108] 34.66 - SID Dataset [13] ReconfigISP [129] 25.65 0.7527 CameraNet [70] 22.47 0.744 S7 ISP [101] Reconfig ISP [129] 23.31 0.7007 HDR+ [37] CameraNet [70] 24.98 0.858 Ciuera and Funt [18] Deep Camera [91] 30.71 - Sony IMX586 Quad Bayer RGB mobile sensor [44] CSANet [44] 23.73 0.8487…”
InfluentialDeep Photo Enhancer: Unpaired Learning for Image Enhancement from Photographs with GANs
Chen, Y., Wang, Y.-C., Kao, M.-H., & Chuang, Y. (2018) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2018.00660 · Cited by 483 · cited 3× · see citing works
“The first compared PyNET with the methods SPADE [85], DPED [45], U-Net [93], Pix2Pix [51], SRGAN [66], VDSR [63], and SRCNN [23].”
Deep joint demosaicking and denoising
Gharbi, M., Chaurasia, G., Paris, S., & Durand, F. (2016) · ACM Transactions on Graphics · doi:10.1145/2980179.2982399 · Cited by 470 · cited 5× · see citing works
“Furthermore, compared with the ADMN [109], CDM [127], Kokkinos [65], Deepjoint 1 [30], and FlexISP [38], the SGNet can remove moiré artifacts more effectively and give images with more definition in high-frequency areas than ADMN and Deepjoint.”
InfluentialImage demosaicing: a systematic survey
Li, X., Gunturk, B. K., & Zhang, L. (2007) · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · doi:10.1117/12.766768 · Cited by 425 · cited 1× · see citing works
“There are some surveys about individual steps of an ISP, such as demosaicing [68] and denoising [25], but not other works about end-to-end ISP via Deep Learning.”
Avoiding Overfitting: A Survey on Regularization Methods for Convolutional Neural Networks
Santos, C. F. G. D., & Papa, J. P. (2022) · ACM Computing Surveys · doi:10.1145/3510413 · Cited by 415 · see citing works
Self-citationBurst Denoising with Kernel Prediction Networks
Mildenhall, B., Barron, J. T., Chen, J., Sharlet, D., Ng, R., & Carroll, R. (2018) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2018.00265 · Cited by 412 · see citing works
CycleISP: Real Image Restoration via Improved Data Synthesis
Zamir, S. W., Arora, A., Khan, S. H., Hayat, M., Khan, F. S., Yang, M., & Shao, L. (2020) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr42600.2020.00277 · Cited by 410 · see citing works
Color image processing pipeline
Ramanath, R., Snyder, W. E., Yoo, Y., & Drew, M. S. (2005) · IEEE Signal Processing Magazine · doi:10.1109/msp.2005.1407713 · Cited by 369 · cited 10× · see citing works
“This strategy would bring many technical issues, related primarily to sensor alignment, difference on light incidence, among others, in addition to an increase in hardware cost [1].”
InfluentialReal-World Super-Resolution via Kernel Estimation and Noise Injection
Ji, X., Cao, Y., Tai, Y. Y., Wang, C., Li, J., & Huang, F. (2020) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw50498.2020.00241 · Cited by 344 · cited 2× · see citing works
“With the SR-RAW dataset, the authors also compared with SR methods, as SRGAN[92], ESRGAN [108], SPSR [109], and RealSR[110].”
Hu, Y., He, H., Xu, C., Wang, B., & Lin, S. C.-F. (2018) · ACM Transactions on Graphics · doi:10.1145/3181974 · Cited by 321 · see citing works
A DEEP LEARNING APPROACH FOR CANCER DETECTION AND RELEVANT GENE IDENTIFICATION
Danaee, P., Ghaeini, R., & Hendrix, D. A. (2016) · Pacific Symposium on Biocomputing (PSB) · doi:10.1142/9789813207813_0022 · Cited by 305 · cited 2× · see citing works
“For that reason, several areas such as Chemistry [5], Medicine [6], Economics [7] and Physics [8], have been able to harness its capacities to accelerate or improve their work, directly impacted by this evolution in the field known as Deep Learning.”
FlexISP: a flexible camera image processing framework
Heide, F., Steinberger, M., Tsai, Y.-T., Rouf, M., Pająk, D., Reddy, D., Gallo, O., Liu, J., Heidrich, W., Egiazarian, K., et al. (2014) · ACM Transactions on Graphics · doi:10.1145/2661229.2661260 · Cited by 298 · see citing works
SURVEY ON IMAGE DENOISING TECHNIQUES
Akhade, K., Ghodekar, S., Kapse, V., Raykar, A., & Wadhvane, S. (2017) · International Journal of Advance Engineering and Research Development · doi:10.21090/ijaerd.99410 · Cited by 271 · see citing works
DeepISP: Toward Learning an End-to-End Image Processing Pipeline
Schwartz, E., Giryes, R., & Bronstein, A. M. (2018) · IEEE Transactions on Image Processing · doi:10.1109/tip.2018.2872858 · Cited by 266 · cited 4× · see citing works
“The results were validated over the SID Dataset [13] and S7 ISP Dataset [102] and showed that this network outperforms the traditional ISP pipelines achieving a higher PSNR value than Camera ISP.”
InfluentialA Survey of Convolutional Neural Networks: Analysis, Applications, and ProspectsPreprint
Li, Z., Yang, W., Peng, S., et al. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2004.02806 · Cited by 261 · cited 26× · see citing works
“PyNET [49] is a novel pyramidal CNN architecture designed to replace the entire ISP pipeline in smartphones.”
InfluentialDistort-and-Recover: Color Enhancement Using Deep Reinforcement Learning
Park, J., Lee, J., Yoo, D., & Kweon, I. S. (2018) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2018.00621 · Cited by 227 · cited 2× · see citing works
“In the second, results were validated over the MIT-Adobe FiveK dataset [11] and compared with HDRNet [31], DPE [16], White-Box [41], Distort-and-Recover [84], and DeepUPE [117] methods where CURL scored better on PSNR and LPIPS metrics, but DeepUPE scored the best on SSIM.”
Machine learning in agricultural and applied economics
Storm, H., Baylis, K., & Heckelei, T. (2019) · European Review of Agricultural Economics · doi:10.1093/erae/jbz033 · Cited by 227 · cited 2× · see citing works
“For that reason, several areas such as Chemistry [5], Medicine [6], Economics [7] and Physics [8], have been able to harness its capacities to accelerate or improve their work, directly impacted by this evolution in the field known as Deep Learning.”
Replacing Mobile Camera ISP with a Single Deep Learning Model
Ignatov, A., Gool, L. V., & Timofte, R. (2020) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw50498.2020.00276 · Cited by 220 · cited 9× · see citing works
“PyNET [31] is a novel pyramidal CNN architecture designed to replace the entire ISP pipeline present in smartphones.”
InfluentialColor image demosaicking: An overview
Menon, D., & Calvagno, G. (2011) · Signal Processing Image Communication · doi:10.1016/j.image.2011.04.003 · Cited by 188 · cited 1× · see citing works
“The article proposes a classification of demosaicing methods in six groups: Edge-sensitive methods, Directional interpolation and decision methods, Frequency domain approaches, Wavelet-based methods, Statistical reconstruction techniques, and Deep learning-based methods [79].”
Hasinoff, S. W. (2014) · Springer eBooks · doi:10.1007/978-0-387-31439-6_482 · Cited by 173 · see citing works
Joint Demosaicing and Denoising With Self Guidance
Liu, L., Jia, X., Liu, J., & Tian, Q. C. (2020) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr42600.2020.00231 · Cited by 127 · cited 17× · see citing works
“In contrast, single-step approaches like SGNet, which focus on specific tasks such as demosaicing and denoising, achieved a PSNR of 26.85 dB while offering better interpretability in noise control.”
InfluentialSupervised Raw Video Denoising With a Benchmark Dataset on Dynamic Scenes
Yue, H., Cao, C., Liao, L., Chu, R., & Yang, J. (2020) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr42600.2020.00237 · Cited by 119 · see citing works
Gait Recognition Based on Deep Learning: A Survey
Filipi Gonçalves dos Santos, C., Oliveira, D. D. S., A. Passos, L., Gonçalves Pires, R., Felipe Silva Santos, D., Pascotti Valem, L., P. Moreira, T., Cleison S. Santana, M., Roder, M., Paulo Papa, J., et al. (2023) · ACM Computing Surveys · doi:10.1145/3490235 · Cited by 108 · see citing works
Self-citationJoint Demosaicing and Denoising via Learned Nonparametric Random Fields
Khashabi, D., Nowozin, S., Jancsary, J., & Fitzgibbon, A. W. (2014) · IEEE Transactions on Image Processing · doi:10.1109/tip.2014.2359774 · Cited by 106 · cited 1× · see citing works
“The results were validated on the MSR demosaicing dataset [125], BSD100 [113], McMaster [83], Urban 100 [84], Kodak [82], and WED [126] datasets and compared with Deepjoint4 [60], Kokkinos [96], Dong [127], DeepISP [70], and DPN [29] methods at three different noise levels (5, 15, 25).”
Schmidhuber, J. (2015) · Scholarpedia · doi:10.4249/scholarpedia.32832 · Cited by 96 · see citing works
CameraNet: A Two-Stage Framework for Effective Camera ISP Learning
Liang, Z., Cai, J., Cao, Z., & Lei Zhang, A. (2021) · IEEE Transactions on Image Processing · doi:10.1109/tip.2021.3051486 · Cited by 94 · see citing works
Gradient based threshold free color filter array interpolation
Pekkucuksen, I., & Altunbasak, Y. (2010) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2010.5654327 · Cited by 92 · cited 1× · see citing works
“The first stage performed the demosaicking by composing the gradient-based threshold-free (GBTF) method [81] and a convolutional neural network to overcome the reduction of image resolution in the subsampling operation.”
Joint demosaicking and denoising by total variation minimization
Condat, L., & Mosaddegh, S. (2012) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2012.6467476 · Cited by 81 · cited 2× · see citing works
“The model was compared with the ADMM [109], Condat [19], Flex-ISP [38], and DemosaicNet 1 [30] on the commonly used benchmark datasets to denoise and demosaicing tasks: Urban 100 [42], Kodak, McMaster [134], and BSD100 [77].”
AIM 2020 Challenge on Learned Image Signal Processing Pipeline
Ignatov, A., Timofte, R., Zhang, Z., et al. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-67070-2_9 · Cited by 76 · cited 10× · see citing works
“By applying this tuned model to two tracks of the Zurich dataset from AIM 2020 Learned Smartphone ISP Challenge [48], the study reached the 5th and 2nd positions, respectively.”
InfluentialLearning Raw Image Denoising With Bayer Pattern Unification and Bayer Preserving Augmentation
Liu, J., Wu, C.-H., Wang, Y., Xu, Q., Zhou, Y., Huang, H., Wang, C., Cai, S., Ding, Y., Fan, H., et al. (2019) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw.2019.00259 · Cited by 71 · see citing works
Co-evolving recurrent neurons learn deep memory POMDPs
Gomez, F. J., & Schmidhuber, J. (2005) · Genetic and Evolutionary Computation Conference (GECCO) · doi:10.1145/1068009.1068092 · Cited by 69 · see citing works
CURL: Neural Curve Layers for Global Image Enhancement
Moran, S., McDonagh, S., & Slabaugh, G. (2021) · Proceedings - International Conference on Pattern Recognition/Proceedings/International Conference on Pattern Recognition · doi:10.1109/icpr48806.2021.9412677 · Cited by 68 · cited 10× · see citing works
“In the first, results were validated on the Samsung S7 dataset [102], and CURL scored the best PSNR and LPIPS metrics when compared to the U-Net [93] and Deep-ISP [102] methods, but tied with the DeepISP method on the SSIM metric.”
InfluentialJoint demosaicing and denoising of noisy bayer images with ADMM
Tan, H., Zeng, X., Lai, S., et al. (2017) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2017.8296823 · Cited by 60 · cited 8× · see citing works
“Furthermore, compared with the ADMN [109], CDM [127], Kokkinos [65], Deepjoint 1 [30], and FlexISP [38], the SGNet can remove moiré artifacts more effectively and give images with more definition in high-frequency areas than ADMN and Deepjoint.”
InfluentialExposure: A White-Box Photo Post-Processing FrameworkPreprint
Hu, Y., He, H., Xu, C., et al. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1709.09602 · Cited by 56 · see citing works
Xia, M., Liu, X., & Wong, T. (2018) · ACM Transactions on Graphics · doi:10.1145/3272127.3275080 · Cited by 52 · cited 1× · see citing works
“The method was compared to Invertible Grayscale [123] and U-net [13] baselines and the results showed better PSNR and SSIM values.”
Learning RAW-to-sRGB Mappings with Inaccurately Aligned Supervision
Zhang, Z., Wang, H., Liu, M., Wang, R., Zhang, J., & Zuo, W. (2021) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv48922.2021.00431 · Cited by 50 · cited 23× · see citing works
“Taking it into account, Zhang et al. [139] proposed a method to train the networks with misaligned images and map RAW to RGB.”
InfluentialLearned Smartphone ISP on Mobile NPUs with Deep Learning, Mobile AI 2021 Challenge: Report
Ignatov, A., Chiang, C.-M., Kuo, H.-K., Sycheva, A., Timofte, R., Chen, M.-H., Lee, M.-Y., Xu, Y.-S., Tseng, Y., Xu, S., et al. (2021) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw53098.2021.00284 · Cited by 44 · cited 6× · see citing works
“” [44] It considers the processing and time of execution in mobile NPUs and the image quality like the other two previous challenges.”
InfluentialComputer Vision – ECCV 2022 Workshops
Karlinsky, L., Michaeli, T., & Nishino, K. (2023) · Lecture notes in computer science · doi:10.1007/978-3-031-25063-7 · Cited by 43 · see citing works
AIM 2019 Challenge on RAW to RGB Mapping: Methods and Results
Ignatov, A., Timofte, R., Ko, S.-J., et al. (2019) · 2019 IEEE/CVF International Conference on Computer Vision Workshop (ICCVW) · doi:10.1109/iccvw.2019.00443 · Cited by 42 · cited 3× · see citing works
“The AIM 2019 Challenge on RAW to RGB Mapping [50] brought new perspectives to this area, with the creation of Zurich Raw to RGB Dataset and the use of comparison metrics which are used in image restoration benchmark datasets.”
InfluentialReconfigISP: Reconfigurable Camera Image Processing Pipeline
Yu, K., Li, Z., Peng, Y., et al. (2021) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv48922.2021.00421 · Cited by 38 · see citing works
A Literature Survey on Blur Detection Algorithms for Digital Imaging
Koik, B. T., & Ibrahim, H. (2013) · International Conference on Artificial Intelligence, Modelling and Simulation (AIMS) · doi:10.1109/aims.2013.50 · Cited by 36 · cited 2× · see citing works
“Given this, many handcraft deblurring algorithms exist in the literature to mitigate this problem[61, 62] and some are included in ISPs.”
PyNet-V2 Mobile: Efficient On-Device Photo Processing With Neural Networks
Ignatov, A., Malivenko, G., Timofte, R., et al. (2022) · Proceedings - International Conference on Pattern Recognition/Proceedings/International Conference on Pattern Recognition · doi:10.1109/icpr56361.2022.9956598 · Cited by 30 · cited 3× · see citing works
“Therefore, ISP Distillation is a step towards achieving similar classification performance on RAW images when compared to RGB. PyNet-V2 Mobile [46] presents a novel approach to on-device photo processing on mobile devices, leveraging the efficacy of neural networks to address the performance…”
InfluentialBala, R., & Mahy, M. (2006) · Color Research & Application · doi:10.1002/col.20239 · Cited by 29 · see citing works
A Survey on Different Image Deblurring Techniques
Vankawala, F., Ganatra, A. P., & Patel, A. V. (2015) · International Journal of Computer Applications · doi:10.5120/20396-2697 · Cited by 28 · cited 1× · see citing works
“Given this, many handcraft deblurring algorithms exist in the literature to mitigate this problem [64, 115] and some are included in ISPs.”
HighEr-Resolution Network for Image Demosaicing and Enhancing
Mei, K., Li, J., Zhang, J., et al. (2019) · doi:10.1109/iccvw.2019.00427 · Cited by 27 · cited 11× · see citing works
“HERNet and LiteISPNet achieve better results in Table 1, but this is due to the use of, respectively, a different data distribution and RAWimages aligned with the ground truth.”
InfluentialDeep Image Demosaicing for Submicron Image Sensors
Kim, I., Song, S., Chang, S., et al. (2019) · Journal of Imaging Science and Technology · doi:10.2352/j.imagingsci.technol.2019.63.6.060410 · Cited by 25 · cited 16× · see citing works
“DPN also implemented a Linear Feature Map Growth, reducing parameter counts when compared to the traditional exponential method, making it more suitable for mobile applications with memory constraints.”
InfluentialThe Shape of Sight: A Homological Framework for Unifying Visual PerceptionPreprint
Li, X. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1802.04723 · Cited by 24 · cited 2× · see citing works
“The results were validated on the MSR demosaicing dataset [125], BSD100 [113], McMaster [83], Urban 100 [84], Kodak [82], and WED [126] datasets and compared with Deepjoint4 [60], Kokkinos [96], Dong [127], DeepISP [70], and DPN [29] methods at three different noise levels (5, 15, 25).”
CSANet: High Speed Channel Spatial Attention Network for Mobile ISP
Hsyu, M.-C., Li, Y., Chen, C.-H., et al. (2021) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw53098.2021.00282 · Cited by 23 · cited 1× · see citing works
“As the second place in the Mobile AI 2021 Learned Smartphone ISP Challenge[39] and first place in PSNR score, the Channel Spatial Attention Network (CSANet)[103] is a network that aims at computational performance and the quality of final image results, inferred at most 90.”
PUCA: Patch-Unshuffle and Channel Attention for Enhanced Self-Supervised Image Denoising
Jang, H., Park, J., Jung, D., et al. (2023) · neural information processing systems · doi:10.52202/075280-0843 · Cited by 22 · see citing works
Askar, A., Askar, A., Pasquato, M., et al. (2019) · Monthly Notices of the Royal Astronomical Society · doi:10.1093/mnras/stz628 · Cited by 21 · cited 2× · see citing works
“For that reason, several areas such as Chemistry [5], Medicine [6], Economics [7] and Physics [8], have been able to harness its capacities to accelerate or improve their work, directly impacted by this evolution in the field known as Deep Learning.”
W-Net: Two-Stage U-Net With Misaligned Data for Raw-to-RGB Mapping
Uhm, K.-H., Kim, S.-W., Ji, S.-W., Cho, S.-J., Hong, J.-P., & Ko, S.-J. (2019) · IEEE International Conference on Computer Vision Workshops · doi:10.1109/iccvw.2019.00448 · Cited by 18 · cited 1× · see citing works
“The ISP-Net used in this work was U-Net with channel attention module [114] and the architecture of CAS-Net was U-Net [93] without channel attention module.”
Image Denoising using Attention-Residual Convolutional Neural Networks
Pires, R. G., Santos, D. F. S., dos Santos, C. F. G., Santana, M. C. S., & Papa, J. P. (2020) · SIBGRAPI Conference on Graphics, Patterns and Images (SIBGRAPI) · doi:10.1109/sibgrapi51738.2020.00022 · Cited by 17 · see citing works
Self-citationColor Filter Array Demosaicking Using Densely Connected Residual Network
Park, B., & Jeong, J. (2019) · IEEE Access · doi:10.1109/access.2019.2939578 · Cited by 16 · see citing works
A Design of Dynamic Defective Pixel Correction for Image Sensor
Yongji, L., & Yuan, X. (2020) · doi:10.1109/icaiis49377.2020.9194921 · Cited by 14 · cited 3× · see citing works
“There are several methods of correction of defective pixels, one of them is to apply the average value of the neighboring points to the pixel identified as defective [50].”
InfluentialMerging-ISP: Multi-exposure High Dynamic Range Image Signal Processing
Chaudhari, P., Schirrmacher, F., Maier, A., et al. (2021) · Lecture notes in computer science · doi:10.1007/978-3-030-92659-5_21 · Cited by 10 · cited 2× · see citing works
“Merging-ISP [44] Approach to reconstructing multiple image layers LDR in just one image HDR.”
Imaeg Blur Reduction for Cell-Phone Cameras Via Adaptive Tonal Correction
Razligh, Q. R., & Kehtarnavaz, N. (2007) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2007.4378904 · Cited by 9 · see citing works
Learning to Sample the Most Useful Training Patches from ImagesPreprint
Sun, S., Chen, L., Slabaugh, G., et al. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2011.12097 · Cited by 7 · cited 6× · see citing works
“In this paper [34], the authors proposed a method based on active learning (data-driven) that learns to select the most suitable patches from an image for the training step without adding additional cost to the inference step.”
InfluentialEEDNet: Enhanced Encoder-Decoder Network for AutoISP
Zhu, Y., Guo, Z., Liang, T., He, X., Li, C., Leng, C., Jiang, B., Zhang, Y., & Cheng, J. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-67070-2_10 · Cited by 7 · cited 3× · see citing works
“In traditional camera pipelines, complex and proprietary hardware processes are used to perform image signal processing [2].”
Dehaze on small-scale datasets via self-supervised learning
Chen, Z., Li, Q., Feng, H., et al. (2023) · The Visual Computer · doi:10.1007/s00371-023-03079-3 · Cited by 7 · see citing works
Multi-Valued and Universal Binary Neurons: Theory, Learning, and Applications [Book Review]
Chen, K., Kvasnička, V., Kanen, P. C., et al. (2001) · IEEE Transactions on Neural Networks · doi:10.1109/tnn.2001.925572 · Cited by 5 · cited 1× · see citing works
“Deep Learning, as a subset of Machine Learning, is comprised of algorithms based on Artificial Neural Networks that use several layers of neurons to extract higher level features from the raw data being provided to it [9, 10, 11].”
Image Compression-Aware Deep Camera ISP Network
Uhm, K.-H., Choi, K., Jung, S., et al. (2021) · IEEE Access · doi:10.1109/access.2021.3116702 · Cited by 4 · cited 2× · see citing works
“ICDC-Net [38] An approach with ISP-Net that addresses JPEG image compression in network training.”
Wilson, P. J. (2015) · Elsevier eBooks · doi:10.1016/b978-0-08-097129-2.00007-6 · Cited by 3 · see citing works
A Large Image Database for Color Constancy Research
Ciurea, F., & Funt, B. (2003) · Color and Imaging Conference · doi:10.2352/cic.2003.11.1.art00028
Joint Denoising and Demosaicking With Green Channel Prior for Real-World Burst Images
Guo, S., Liang, Z., & Zhang, L. (2021) · IEEE Transactions on Image Processing · doi:10.1109/tip.2021.3100312
Flagged first, then by citations
Research works printed but not linked: 32
- S. M. A. Sharif, Rizwan Ali Naqvi, and Mithun Biswas. 2021. Beyond joint demosaicking and denoising: An image processing pipeline for a pixel-bin image sensor. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 233–242.
- Forrest Iandola Peter H. Jin Kurt Keutzer Bichen Wu and Alvin Wan. 2016. SqueezeDet: Unified small low power fully convolutional neural networks for real-time object detection for autonomous driving. arxiv:cs.CV/1612.01051
- Liang-Chieh Chen George Papandreou Iasonas Kokkinos Kevin Murphy and Alan L. Yuille. 2017. DeepLab: Semantic image segmentation with deep convolutional nets atrous convolution and fully connected CRFs. arxiv:cs.CV/1606.00915
- Linhui Dai Xiaohong Liu Chengqi Li and Jun Chen. 2020. AWNet: Attentive wavelet network for image ISP. arxiv:eess.IV/2008.09228
- Chao Dong Chen Change Loy Kaiming He and Xiaoou Tang. 2015. Image super-resolution using deep convolutional networks. arxiv:cs.CV/1501.00092
- Zhihao Fan, Xun Wu, Fanqing Meng, Yaqi Wu, and Feng Zhang. 2023. OTST: A two-phase framework for joint denoising and remosaicing in RGBW CFA. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2832–2841.
- Yu Guo Qiyu Jin Gabriele Facciolo Tieyong Zeng and Jean-Michel Morel. 2020. Residual learning for effective joint demosaicing-denoising. arxiv:cs.CV/2009.06205
- Saumya Gupta Diplav Srivastava Umang Chaturvedi Anurag Jain and Gaurav Khandelwal. 2021. Del-Net: A single-stage network for mobile camera ISP. arxiv:eess.IV/2108.01623
- Geoffrey Hinton Oriol Vinyals and Jeff Dean. 2015. Distilling the knowledge in a neural network. arxiv:stat.ML/1503.02531
- Andrey Ignatov, Anastasia Sycheva, Radu Timofte, Yu Tseng, Yu-Syuan Xu, Po-Hsiang Yu, Cheng-Ming Chiang, Hsien-Kai Kuo, Min-Hung Chen, Chia-Ming Cheng, et al. 2022. MicroISP: Processing 32mp photos on mobile devices with deep learning. In European Conference on Computer Vision. Springer, 729–746.
- Jun Nishimura Timo Gerasimow Sushma Rao Aleksandar Sutic Chyuan-Tyng Wu and Gilad Michael. 2019. Automatic ISP image quality tuning using non-linear optimization. arxiv:cs.CV/1902.09023
- S. Ren K. He X. Zhang and J. Sun. 2016. Deep residual learning for image recognition. arxiv:cs.CV/1512.03385
- Hakki Can Karaimer and Michael S. Brown. 2016. A software platform for manipulating the camera imaging pipeline. In European Conference on Computer Vision (ECCV’16).
- Byung-Hoon Kim, Joonyoung Song, Jong Chul Ye, and JaeHyun Baek. 2020. PyNET-CA: Enhanced PyNET with channel attention for end-to-end mobile image signal processing. In European Conference on Computer Vision. Springer, 202–212.
- Filippos Kokkinos and Stamatios Lefkimmiatis. 2018. Deep image demosaicking using a cascade of convolutional residual denoising networks. In Proceedings of the European Conference on Computer Vision (ECCV’18). 303–319.
- Juncheng Li, Faming Fang, Kangfu Mei, and Guixu Zhang. 2018. Multi-scale residual network for image super-resolution. In ECCV.
- Cheng Ma Yongming Rao Yean Cheng Ce Chen Jiwen Lu and Jie Zhou. 2020. Structure-preserving super resolution with gradient guidance. arxiv:eess.IV/2003.13081
- Aamir Mustafa Aliaksei Mikhailiuk Dan Andrei Iliescu Varun Babbar and Rafal K. Mantiuk. 2021. Training a task-specific image reconstruction loss. arxiv:eess.IV/2103.14616
- Guocheng Qian Yuanhao Wang Chao Dong Jimmy S. Ren Wolfgang Heidrich Bernard Ghanem and Jinjin Gu. 2021. Rethinking the pipeline of demosaicing denoising and super-resolution. arxiv:eess.IV/1905.02538
- Ramchalam Ramakrishnan, Shangling Jui, and Vahid Partovi Nia. 2019. Deep demosaicing for edge implementation. In International Conference on Image Analysis and Recognition. Springer, 275–286.
- Eli Schwartz Alex Bronstein and Raja Giryes. 2021. ISP distillation. arxiv:cs.CV/2101.10203
- 10.1109/tip.2018.2872858 · DOI
- Deqing Sun Xiaodong Yang Ming-Yu Liu and Jan Kautz. 2018. PWC-Net: CNNs for optical flow using pyramid warping and cost volume. arxiv:cs.CV/1709.02371
- Stefan Roth and Tobias Plötz. 2017. Benchmarking denoising algorithms with real photographs. arxiv:cs.CV/1707.01313
- Sanghyun Woo Jongchan Park Joon-Young Lee and In So Kweon. 2018. CBAM: Convolutional block attention module. arxiv:cs.CV/1807.06521
- Chyuan-Tyng Wu, Leo F. Isikdogan, Sushma Rao, Bhavin Nayak, Timo Gerasimow, Aleksandar Sutic, Liron Ain-kedem, and Gilad Michael. 2019. VisionISP: Repurposing the image signal processor for computer vision applications. In 2019 IEEE International Conference on Image Processing (ICIP’19). IEEE, 4624–4628.
- Yazhou Xing Zian Qian and Qifeng Chen. 2021. Invertible image signal processing. arxiv:eess.IV/2103.15061
- Ke Yu Chao Dong Chen Change Loy Xintao Wang Kelvin C.K. Chan. 2019. EDVR: Video restoration with enhanced deformable convolutional networks. arxiv:cs.CV/1905.02716
- Bing Xu Naiyan Wang Tianqi Chen and Mu Li. 2015. Empirical evaluation of rectified activations in convolutional network. arxiv:cs.LG/1505.00853
- Georgi Zapryanov, Ivanova, and Iva Nikolova. 2012. Automatic white balance algorithms for digital still cameras - a comparative study. Information Technologies and Control 1 (012012), 16–22.
- Molin Zhang, Junshen Xu, Yamin Arefeen, and Elfar Adalsteinsson. 2024. Zero-shot self-supervised joint temporal image and sensitivity map reconstruction via linear latent space. In Medical Imaging with Deep Learning. PMLR, 1713–1725.
- Xuaner Cecilia Zhang Qifeng Chen Ren Ng and Vladlen Koltun. 2019. Zoom To learn learn To Zoom. arxiv:cs.CV/1905.05169
Cited by
Sorted by citations
Multi-modal anomaly detection for microservice system through nested graph diffusion reconstruction
Fan, M., Zhang, X., Wang, P., & Cao, Z. (2025) · Applied Intelligence · doi:10.1007/s10489-025-06681-1 · Cited by 7 · see citing works
Learning Deep ISP for High-Speed Cameras: Achieving DSLR-Quality Imaging Under High Frame Rates
Chen, H., Liu, T., Tu, T., Wei, W., Jin, Y., & Chen, E. (2025) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2025.3637771 · Cited by 2 · see citing works
“Broadly, these approaches can be categorized into traditional prior-based methods [9], [10] and data-driven methods [6], [8], [11], [19], [20], [21], [22], [23].”
Deep Learning-Enabled Cross-Device Calibration for Smartphone-Based Optical Biosensing: A Review
Zuo, H., Geng, Q., Li, H., Tao, M., Feng, Y., & Geng, Z. (2026) · IEEE Sensors Journal · doi:10.1109/jsen.2026.3716330 · Cited by 1 · see citing works
“Demosaicing, denoising, sharpening, color matrices, gamma correction, tone mapping, and compression alter the mapping between pixel values and radiant intensity, making measurements nonexchangeable across manufacturers and capture modes [24].”
Lightweight RAW Object Detection for Automated Driving
Souvalioti, G., Goswami, A., Singh, A. R., et al. (2025) · IEEE Intelligent Vehicles Symposium · doi:10.1109/iv64158.2025.11097580
“This problem can be rectified by deep learning counterparts, though at the cost of increased complexity [9]–[11].”
Human motion posture recognition and analysis based on video image data and PSO-BP
Yuan, Z. (2025) · Journal of Computational Methods in Sciences and Engineering · doi:10.1177/14727978251361856
High dynamic camera design for night use combined with AI-ISP
Qin, Z., Tan, Q., & Peng, Z. (2025) · The Computer Journal · doi:10.1093/comjnl/bxaf128
Yang, T., Guo, L., Li, J., et al. (2026) · Frontiers in Neurorobotics · doi:10.3389/fnbot.2026.1866481
“…deep learning techniques have substantially improved ISP performance by training convolutional neural networks (CNNs) on large scale datasets, enabling accurate reconstruction of color, illumination, texture, and high dynamic range details from RAW image data (Santos et al., 2025; Li et al., 2024).”
A bio-inspired framework for lightweight RAW-to-sRGB mapping and color optimization
Zhao, P., Chen, Z., Yang, F., et al. (2026) · Pattern Recognition · doi:10.1016/j.patcog.2026.114761
P‐5.2: A Two‐Stage Edge‐Aware Adaptive Anti‐Aliasing and Anti‐Artifact Demosaicing Algorithm
Shi, Y., Ran, S., & Yan, L. (2026) · SID Symposium Digest of Technical Papers · doi:10.1002/sdtp.70003_348
- Authors
- Santos, C. F. G. D., Arrais, R. R., Silva, J. V. S. D., Silva, M. H. M. D., Neto, W. B. G. D. A., Lopes, L. T., Bileki, G. A., Lima, I. O., Rondon, L. B., Souza, B. M. D., Regazio, M. C., Dalapicola, R. C., & Tasca, A. A.
- Affiliations
- Computer Science, UFSCar, Sao Carlos, Brazil and Eldorado Institute, Campinas, Brazil
- Instituto de Pesquisas Eldorado (BR)
- Eldorado Institute, Campinas, Brazil
- ORCID
- Santos, Claudio Filipi Goncalves dos
- Arrais, Rodrigo Reis
- Silva, Jhessica Victoria Santos da
- Silva, Matheus Henrique Marques da
- Neto, Wladimir Barroso Guedes de Araujo
- Lopes, Leonardo Tadeu
- Bileki, Guilherme Augusto
- Lima, Iago Oliveira
- Rondon, Lucas Borges
- Souza, Bruno Melo de
- Regazio, Mayara Costa
- Dalapicola, Rodolfo Coelho
- Tasca, Arthur Alves
- Venue
- ACM Computing Surveys
- Published
- 2025-01-22
- Volume
- 57
- Issue
- 5
- Pages
- 1-44
- ISSN
- 0360-0300 (ISSN-L)
- 1557-7341
- OpenAlex
- W4405580099
- Semantic Scholar
- 808a2f197901454f367e3f23151307b486c72015
- Cited by
- 9
- References
- 102
- Citations by year
- 2026: 4 · 2025: 5
- FWCI
- 1.27
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- Locations
- ACM Computing Surveys (publishedVersion, cc-by)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Image Processing Techniques
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Image Enhancement Techniques
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Keywords
- image signal processing, deep learning, camera image processing, color filter array, demosaicing, image denoising, image enhancement, hardware-software comparison
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
