Xception: Deep Learning with Depthwise Separable Convolutions
The DOI matches its Crossref record; 20 of the 21 references it cites were found in a registry or an index.
Abstract
We present an interpretation of Inception modules in convolutional neural networks as being an intermediate step in-between regular convolution and the depthwise separable convolution operation (a depthwise convolution followed by a pointwise convolution). In this light, a depthwise separable convolution can be understood as an Inception module with a maximally large number of towers. This observation leads us to propose a novel deep convolutional neural network architecture inspired by Inception, where Inception modules have been replaced with depthwise separable convolutions. We show that this architecture, dubbed Xception, slightly outperforms Inception V3 on the ImageNet dataset (which Inception V3 was designed for), and significantly outperforms Inception V3 on a larger image classification dataset comprising 350 million images and 17,000 classes. Since the Xception architecture has the same number of parameters as Inception V3, the performance gains are not due to increased capacity but rather to a more efficient use of model parameters.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“The top-5 popular baselines adopted in the evaluation are XceptionNet (Chollet 2017), Afchar et al. (2018), Nguyen et al. (2019a), ResNet (He et al. 2016), and Yang et al. (2019b).”
From Countering Malicious DeepFakes: Survey, Battleground, and Horizon (2022) · cites this paper as Chollet 2017
In the citing paper
“The main idea behind Xception is its depthwise separable convolution [125].”
From Review of deep learning: concepts, CNN architectures, challenges, applications, future directions (2021) · cites this paper as [125]
In the citing paper
“Depthwise separable convolution (DSC) has been found in past research to have a smaller number of parameters than traditional convolution [26].”
From Multimodal Emotion Recognition Based on Facial Expressions, Speech, and EEG (2024) · cites this paper as [26]
In the citing paper
“In the research [24], a depthwise separable convolution neural network on FPGA is being reviewed due to its ability to reduce the number of model parameters and improve calculation speeds, which makes it widely used in mobile edge computing applications such as Xception [25] and MobileNet [26].”
From Embedded Intelligence on FPGA: Survey, Applications and Challenges (2021) · cites this paper as [25]
In the citing paper
“For classification (Table 10), the performance of Xception [12], LipForensics [21], Face X-Ray [36], and SBI [56] also drops compared to LAV-DF.”
From AV-Deepfake1M: A Large-Scale LLM-Driven Audio-Visual Deepfake Dataset (2024) · cites this paper as [12]
5 more claims about this paper
Create a free account to see every claim other papers make about this one.
Citations
Cites
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.90 · Cited by 235,474 · cited 2×
“in [4], which our proposed architecture uses extensively.”
BackgroundVery Deep Convolutional Networks for Large-Scale Image Recognition
Simonyan, K., & Zisserman, A. (2014) · arXiv · doi:10.48550/arxiv.1409.1556 · Cited by 114,148
ImageNet classification with deep convolutional neural networks
Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2017) · Communications of the ACM · doi:10.1145/3065386 · Cited by 112,155 · cited 2×
“In 2012, these ideas were refined into the AlexNet architecture [9], where convolution operations were being repeated multiple times in-between max-pooling operations, allowing the network to learn richer features at every spatial scale.”
BackgroundGoing deeper with convolutions
Szegedy, C., Liu, W., Jia, Y., Sermanet, P., Reed, S., Anguelov, D., Erhan, D., Vanhoucke, V., & Rabinovich, A. (2015) · 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2015.7298594 · Cited by 48,562
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×
“Since its first introduction, Inception has been one of the best performing family of models on the ImageNet dataset [14], as well as internal datasets in use at Google, in particular JFT [5].”
BackgroundMethodRethinking the Inception Architecture for Computer Vision
Szegedy, C., Vanhoucke, V., Ioffe, S., Shlens, J., & Wojna, Z. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.308 · Cited by 32,138
Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
Ioffe, S., & Szegedy, C. (2015) · arXiv · doi:10.48550/arxiv.1502.03167 · Cited by 27,818
MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications
Howard, A. G., Zhu, M., Chen, B., Kalenichenko, D., Wang, W., Weyand, T., Andreetto, M., & Adam, H. (2017) · arXiv · doi:10.48550/arxiv.1704.04861 · Cited by 19,015
Visualizing and understanding convolutional networks
Zeiler, M. D., & Fergus, R. (2014) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-10590-1_53 · Cited by 15,839 · cited 2×
“What followed was a trend to make this style of network increasingly deeper, mostly driven by the yearly ILSVRC competition; first with Zeiler and Fergus in 2013 [25] and then with the VGG architecture in 2014 [18].”
BackgroundDistilling the Knowledge in a Neural Network
Hinton, G., Vinyals, O., & Dean, J. (2015) · arXiv · doi:10.48550/arxiv.1503.02531 · Cited by 14,204 · cited 2×
“Since its first introduction, Inception has been one of the best performing family of models on the ImageNet dataset [14], as well as internal datasets in use at Google, in particular JFT [5].”
Background
Research works printed but not linked: 7
- lin, 2013, Network in Network
- sifre, 2014, Rigid-motion scattering for image classification
- silberman, 2016, Tf-slim
- chollet, 2015, Keras
- abadi, 2015, Tensor-Flow Large-scale machine learning on heterogeneous systems
- tieleman, 0, Divide the gradient by a running average of its recent magnitude, COURSERA Neural Networks for Machine Learning, 4
- vanhoucke, 2014, Learning visual representations at scale, ICLRE
Cited by
Items 1,201–1,227 of 1,227
Barreto, A. G., de Oliveira, J. M., Gois, F. N. B., Cortez, P. C., & de Albuquerque, V. H. C. (2023) · Bioengineering · doi:10.3390/bioengineering10091098 · Cited by 8
RefineTAD: Learning Proposal-free Refinement for Temporal Action Detection
Feng, Y., Zhang, Z., Quan, R., Wang, L., & Qin, J. (2023) · Proceedings of the 31st ACM International Conference on Multimedia · doi:10.1145/3581783.3611872 · Cited by 8
Atamna, M., Tkachenko, I., & Miguet, S. (2023) · 2023 IEEE International Conference on Image Processing (ICIP) · doi:10.1109/icip49359.2023.10222043 · Cited by 8
“In this paper, we focus on training and evaluating the generalization performance of two types of architectures: XceptionNet [13], a popular CNN architecture for deepfake image classification which, despite its age, is still widely used for benchmarking purposes, and a spatio-temporal architecture…”
Learning Versatile Convolution Filters for Efficient Visual Recognition
Han, K., Wang, Y., Xu, C., Xu, C., Wu, E., & Tao, D. (2022) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2021.3114368 · Cited by 8
“Recently, a series of efficient operations are proposed for building lightweight neural networks [8], [16],”
BackgroundMethodInfluentialMultimodal deep learning ensemble framework for skin cancer detection
Saeed, M. A., Afify, Y. M., Badr, N. L., & Helal, N. A. (2025) · Scientific Reports · doi:10.1038/s41598-025-30534-z · Cited by 7
“Xception Xception 50 , short for "Extreme Inception," is a deep learning model introduced by Google.”
Adversarially Robust Deepfake Video Detection
Devasthale, A., & Sural, S. (2022) · 2022 IEEE Symposium Series on Computational Intelligence (SSCI) · doi:10.1109/ssci51031.2022.10022079 · Cited by 7
“Likewise, XceptionNet [14] is another CNN architecture that has been shown to achieve high accuracy for the deepfake detection task.”
MethodPredicting bulge to total luminosity ratio of galaxies using deep learning
Grover, H., Bait, O., Wadadekar, Y., & Mishra, P. K. (2021) · Monthly Notices of the Royal Astronomical Society · doi:10.1093/mnras/stab1935 · Cited by 7
Tong, L., Corrigan, A., Kumar, N. R., Hallbrook, K., Orme, J., Wang, Y., & Zhou, H. (2022) · Scientific Reports · doi:10.1038/s41598-022-12099-3 · Cited by 6
Towards lightweight deep neural network for smart agriculture on embedded systems
Du, P., Polonelli, T., Magno, M., & Cheng, Z. (2022) · 2022 IEEE Sensors Applications Symposium (SAS) · doi:10.1109/sas54819.2022.9881382 · Cited by 5
King, O. N. F., Levik, K. E., Sandy, J., & Basham, M. (2024) · Acta Crystallographica Section D Structural Biology · doi:10.1107/s2059798324009276 · Cited by 5
Rami Naidji, M., & Elberrichi, Z. (2024) · International Journal of Computing and Digital Systems · doi:10.12785/ijcds/1501102 · Cited by 4
Luo, J., Mao, Q., Wang, Y., Shi, Z., & Hei, X. (2022) · Brain Science Advances · doi:10.26599/bsa.2022.9050011 · Cited by 4
“SConv indicates separable convolution consisting of a depthwise convolution and a pointwise convolution [17, 26].”
BackgroundChrominance and Luminance: a study to detect deepfakes
Talib, M. A., Nasir, Q., Nassif, A. B., Fadhl, N. B., & Gouda, O. M. (2025) · Multimedia Tools and Applications · doi:10.1007/s11042-025-20618-3 · Cited by 5
“The modular formation of the NasNet convolutional cell is connected to the preceding methods on ImageNet such as VGG [56], Inception [57–59], ResNet/ResNext [24, 60], and Xception/MobileNet [61, 62].”
Basereh, M., Abikenari, M. A., Sadeghzadeh, S., Dunn, T., Freichel, R., Siddarth, P., Ghahremani, D., Lavretsky, H., & Buch, V. P. (2025) · Diagnostics · doi:10.3390/diagnostics15232997 · Cited by 4
“This architecture offers significant computational savings while retaining competitive performance [30].”
SIGMA-DF: Single-Side Guided Meta-Learning for Deepfake Detection
Han, B., Li, J., Ren, W., Luo, M., Liu, J., & Cao, X. (2023) · Proceedings of the 2023 ACM International Conference on Multimedia Retrieval · doi:10.1145/3591106.3592282 · Cited by 4
“Training Set FF++ Test Set Venue Xception[7] Frame FF++ 99.1 CVPR2017 F 3 -Net[33] Frame FF++ 98.1 ECCV2020 Two-Branch[28] Frame FF++ 98.7 ECCV2020 Multi-Attention[43] Frame FF++ 99.3 CVPR2021 GFF[27] Frame FF++ 98.7 CVPR2021 Local Relation[6] Frame FF++ 99.5 AAAI2021 LTW[36] Frame FF++ 99.2 AAAI2021 DCL[37] Frame FF+…”
BackgroundInfluentialVerenich, E., Martin, T., Velasquez, A., Khan, N., & Hussain, F. (2021) · arXiv · doi:10.48550/arxiv.2109.00573 · Cited by 4
“[17] proposed CoroNet, a network based on the Xception [18] architecture for diagnosis of COVID-19 from chest X-rays.”
MethodAV-Deepfake1M++: A Large-Scale Audio-Visual Deepfake Benchmark with Real-World Perturbations
Cai, Z., Kuckreja, K., Ghosh, S., Chuchra, A., Khan, M. H., Tariq, U., Gedeon, T., & Dhall, A. (2025) · Proceedings of the 33rd ACM International Conference on Multimedia · doi:10.1145/3746027.3761979 · Cited by 4
“In this section, we report the benchmark of several baseline methods [4, 5, 9] and top teams.”
InfluentialDeepfake Detection Using Transfer Learning
Khan, S. A., & Valles, D. (2024) · 2024 IEEE 15th Annual Ubiquitous Computing, Electronics & Mobile Communication Conference (UEMCON) · doi:10.1109/uemcon62879.2024.10754706 · Cited by 2
Atamna, M., Tkachenko, I., & Miguet, S. (2025) · Lecture Notes in Computer Science · doi:10.1007/978-3-031-88223-4_16 · Cited by 1
“For image classification, we use Xcep-tionNet [6], a popular CNN architecture for image classification which is widely used for benchmarking purposes in deepfake detection [27].”
Baz, M. (2024) · Mathematics · doi:10.3390/math12213320 · Cited by 1
Deepfake Detection with Data Privacy Protection
Wu, M., Wang, F., Wu, X., Yu, F., Wang, B., & Song, Z. (2022) · 2022 IEEE 24th International Workshop on Multimedia Signal Processing (MMSP) · doi:10.1109/mmsp55362.2022.9949458
“The shuffle-block is a residual structure, and composed of single kernel group convolution, channel shuffle and DWConv [23].”
MoCHA-former: Moiré-Conditioned Hybrid Adaptive Transformer for Video Demoiréing
Sung, J., Roh, C., Eom, C., & Oh, J. (2026) · Neurocomputing · doi:10.1016/j.neucom.2025.132477
High-resolution mri guided whole mouse brain neuronal cell type atlas using deep learning
Han, X., Hu, R., Liu, Z., Chen, J., Jafry, M., Song, H., Zhao, Y., Lin, M., White, L. E., Johnson, G. A., et al. (2026) · Communications Biology · doi:10.1038/s42003-026-10608-y
“Outputs are concatenated, re fi ned with depthwise-separable and pointwise convolutions 78 , and modulated by a channel-attention mechanism to enhance discriminative features 79 .”
Anand, V., & Rodriguez, I. F. (2026) · 2026 IEEE 5th International Conference on AI in Cybersecurity (ICAIC) · doi:10.1109/icaic67076.2026.11395676
Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, L., & Polosukhin, I. (2017) · arXiv · doi:10.48550/arxiv.1706.03762 · Cited by 120,404
“Separable convolutions [6], however, decrease the complexity considerably, to O(k · n · d + n · d(2)).”
BackgroundForensicTransfer: Weakly-supervised Domain Adaptation for Forgery Detection
Cozzolino, D., Thies, J., Rössler, A., Riess, C., Nießner, M., & Verdoliva, L. (2018) · arXiv · doi:10.48550/arxiv.1812.02510 · Cited by 221
“XceptionNet [10]: is based on depth-wise separable convolution layers with residual connections.”
BackgroundMethodInfluentialConstructing Fast Network through Deconstruction of Convolution
Jeon, Y., & Kim, J. (2018) · arXiv · doi:10.48550/arxiv.1806.07370
“Recently, depthwise separable convolution has been shown to achieve similar accuracy to naive convolution while reducing computational complexity[2, 8, 18].”
Background
Metadata
- Authors
- Chollet, F.
- Affiliations
- Google (United States) (US)
- Corresponding author
- Chollet, Francois
- Venue
- 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
- Publisher
- IEEE
- Published
- 2017-07
- Pages
- 1800-1807
- ISSN
- 1063-6919 (ISSN-L)
- OpenAlex
- W2531409750
- Cited by
- 20,380
- References
- 21
- Citations by year
- 2026: 1618 · 2025: 3098 · 2024: 3287 · 2023: 3281 · 2022: 3003 · 2021: 2735 · 2020: 1859 · 2019: 1096 · 2018: 356 · 2017: 42
- FWCI
- 268.33
- Type
- conference paper
- Language
- en
- Open access
- closed
- Locations
- IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings (publishedVersion)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Physical Sciences › Computer Science › Artificial Intelligence › Domain Adaptation and Few-Shot Learning
- Physical Sciences › Computer Science › Artificial Intelligence › Machine Learning and Data Classification
- Keywords
- Xception, depthwise separable convolution, Inception module, InceptionV3, convolutional neural network architecture, image classification, parameter efficiency, ImageNet, large-scale image classification
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
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.