The DOI matches its Crossref record; 31 of the 32 references it cites were found in a registry or an index.
Deep Learning for Emotion Recognition on Small Datasets using Transfer Learning
Abstract
This paper presents the techniques employed in our team's submissions to the 2015 Emotion Recognition in the Wild contest, for the sub-challenge of Static Facial Expression Recognition in the Wild. The objective of this sub-challenge is to classify the emotions expressed by the primary human subject in static images extracted from movies. We follow a transfer learning approach for deep Convolutional Neural Network (CNN) architectures. Starting from a network pre-trained on the generic ImageNet dataset, we perform supervised fine-tuning on the network in a two-stage process, first on datasets relevant to facial expressions, followed by the contest's dataset. Experimental results show that this cascading fine-tuning approach achieves better results, compared to a single stage fine-tuning with the combined datasets. Our best submission exhibited an overall accuracy of 48.5% in the validation set and 55.6% in the test set, which compares favorably to the respective 35.96% and 39.13% of the challenge baseline.
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
“Since the number of labeled examples for FER remains relatively limited, large 11 models are likely to overfit the data [106], [107], [89], [108].”
From Deep learning for human affect recognition: Insights and new developments (2021) · cites this paper as [107]
Citations
Cites
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
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
Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation
Girshick, R., Donahue, J., Darrell, T., & Malik, J. (2014) · 2014 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2014.81 · Cited by 32,684
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
Robust Real-Time Face Detection
Viola, P., & Jones, M. J. (2004) · International Journal of Computer Vision · doi:10.1023/b:visi.0000013087.49260.fb · Cited by 14,097
Advances in Neural Information Processing Systems 25
Zhang, Y., Sutton, C. A., Storkey, A., & Zoubin Ghahramani, A. (2012) · Cited by 12,653
How transferable are features in deep neural networks?
Yosinski, J., Clune, J., Bengio, Y., & Lipson, H. (2014) · arXiv · doi:10.48550/arxiv.1411.1792 · Cited by 6,061
DeCAF: A Deep Convolutional Activation Feature for Generic Visual Recognition
Donahue, J., Jia, Y., Vinyals, O., Hoffman, J., Zhang, N., Tzeng, E., & Darrell, T. (2013) · arXiv · doi:10.48550/arxiv.1310.1531 · Cited by 4,019
Return of the Devil in the Details: Delving Deep into Convolutional Nets
Chatfield, K., Simonyan, K., Vedaldi, A., & Zisserman, A. (2014) · Proceedings of the British Machine Vision Conference 2014 · doi:10.5244/c.28.6 · Cited by 3,335
Recognizing action units for facial expression analysis
Tian, Y.-I., Kanade, T., & Cohn, J. F. (2001) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/34.908962 · Cited by 1,745
Research works printed but not linked: 1
- J. R. S. Widen and A. Brooks . Anger and disgust: Discrete or overlapping categories ? In Proceedings of the 2004 APS Annual Convention , 2004 . J. R. S. Widen and A. Brooks. Anger and disgust: Discrete or overlapping categories? In Proceedings of the 2004 APS Annual Convention, 2004.
Cited by
Sorted by citations
Items 201–300 of 655
EvilModel 2.0: Bringing Neural Network Models into Malware Attacks
Wang, Z., Liu, C., Cui, X., Yin, J., & Wang, X. (2022) · Computers & Security · doi:10.1016/j.cose.2022.102807 · Cited by 23
Detection of Skin Cancer using Artificial Intelligence & Machine Learning Concepts
Suhasini, V. K., Patil, P. B., Vijaykumar, K. N., Manjunatha, S. C., Sudha, T., Gowda, K. P., Ramachandraiah, G., Pavithra, G., & Manjunath, T. C. (2022) · IEEE International Conference on Cybernetics, Cognition and Machine Learning Applications (ICCCMLA) · doi:10.1109/icccmla56841.2022.9989146 · Cited by 23
“1, is relatively similar to the one developed by Aya Abu Ali and Hasan Al-Marzouqi [11][12].”
BackgroundResultInfluentialBoosting-POOF: Boosting Part Based One vs One Feature for Facial Expression Recognition in the Wild
Liu, Z., Li, S., & Deng, W. (2017) · International Conference on Automatic Face and Gesture Recognition · doi:10.1109/fg.2017.120 · Cited by 23
JellyNet: The convolutional neural network jellyfish bloom detector
Mcilwaine, B., & Casado, M. R. (2021) · International Journal of Applied Earth Observation and Geoinformation · doi:10.1016/j.jag.2020.102279 · Cited by 24
“Transfer learning is crucial to achieve greater than 90% test accuracy for the comparatively small (4000 inputs) jellyfish dataset (Ng et al., 2015).”
BackgroundOn line emotion detection using retrainable deep neural networks
Kollias, D., Tagaris, A., & Stafylopatis, A. (2016) · IEEE Symposium Series on Computational Intelligence (SSCI) · doi:10.1109/ssci.2016.7850049 · Cited by 22
“In the above-mentioned Challenges, the winner methods have used networks pre-trained with the FER-2013 and the Toronto Face Dataset [25], [26], [27].”
MethodPooling Map Adaptation in Convolutional Neural Network for Facial Expression Recognition
Li, Z., Han, S., Khan, A. S., Cai, J., Meng, Z., O’Reilly, J. D., & Tong, Y. (2019) · IEEE International Conference on Multimedia and Expo (ICME) · doi:10.1109/icme.2019.00194 · Cited by 22
“[26] Single 48.”
An enhanced fitness function to recognize unbalanced human emotions data
Acharya, D., Varshney, N., Vedant, A., Saxena, Y., Tomar, P., Goel, S., & Bhardwaj, A. (2020) · Expert Systems with Applications · doi:10.1016/j.eswa.2020.114011 · Cited by 22
A Novel Deep Learning Model to Secure Internet of Things in Healthcare
Ahmad, U., Song, H., Bilal, A., Mahmood, S., Alazab, M., Jolfaei, A., Ullah, A., & Saeed, U. (2020) · Studies in computational intelligence · doi:10.1007/978-3-030-57024-8_15 · Cited by 22
Towards scalable and reusable predictive models for cyber twins in manufacturing systems
Giannetti, C., & Essien, A. E. (2021) · Journal of Intelligent Manufacturing · doi:10.1007/s10845-021-01804-0 · Cited by 24
“For instance, if the target dataset is small, and the model has a large number of parameters, fine-tuning many layers may result in over-fitting (Ng et al. 2015).”
Cross-scenario automatic sleep stage classification using transfer learning and single-channel EEG
He, Z., Tang, M., Wang, P., Du, L., Chen, X., Cheng, G., & Fang, Z. (2022) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2022.104501 · Cited by 22
Machine Learning for Detecting Total Knee Arthroplasty Implant Loosening on Plain Radiographs
Kim, M. S., Cho, R.-K., Yang, S.-C., Hur, J. H., & In, Y. (2023) · Bioengineering · doi:10.3390/bioengineering10060632 · Cited by 22
“The method proposed to solve this problem is a transfer learning model, which increases the accuracy of the model by adding training data suitable for the study to an existing neural network trained with large amounts of more general data [52].”
MethodImproving Speech Related Facial Action Unit Recognition by Audiovisual Information Fusion
Meng, Z., Han, S., Liu, P., & Tong, Y. (2018) · IEEE Transactions on Cybernetics · doi:10.1109/tcyb.2018.2840090 · Cited by 22
A Weakly Supervised learning technique for classifying facial expressions
Happy, S. L., Dantcheva, A., & Brémond, F. (2019) · Pattern Recognition Letters · doi:10.1016/j.patrec.2019.08.025 · Cited by 22
“For example, improved performance was observed in emotion classification (Ng et al., 2015) using the transfer of network weights trained on ImageNet.”
MethodArora, M., Mangipudi, P., & Dutta, M. K. (2021) · Journal of Ambient Intelligence and Humanized Computing · doi:10.1007/s12652-020-02867-2 · Cited by 24
“Transfer learning approach inherent several fascinating advantages like efficient handling of larger dataset (Ng et al. 2015) and automatic transferring of most robust and discriminatory features (Kamilaris and Prenafeta-Boldú 2018) for image analysis.”
BackgroundFighting Covid-19 through fast delivery of a modular quarantine camp with smart construction
Zhang, Z., Pan, W., & Zheng, Z. (2020) · Proceedings of the Institution of Civil Engineers - Civil Engineering · doi:10.1680/jcien.20.00025 · Cited by 21
Discriminative attention-augmented feature learning for facial expression recognition in the wild
Zhou, L., Fan, X., Tjahjadi, T., & Choudhury, S. D. (2021) · Neural Computing and Applications · doi:10.1007/s00521-021-06045-z · Cited by 21
Emotion Identification in Movies through Facial Expression Recognition
Almeida, J. E., Vilaça, L., Teixeira, I. N., & Viana, P. (2021) · Applied Sciences · doi:10.3390/app11156827 · Cited by 21
“This issue is commonly tackled by using multistage fine-tuning strategies based on different combinations of the training dataset to enhance their performance [54] or by using facial recognition feature extractors and regularizing them with facial expression information [55].”
BackgroundRen, C., Chen, J., Li, R., Zheng, W., Chen, Y., Yang, Y., Zhang, X., & Hu, B. (2024) · Knowledge-Based Systems · doi:10.1016/j.knosys.2024.112669 · Cited by 21
Meta transfer learning for facial emotion recognition
Nguyen, D., Nguyen, K., Sridharan, S., Abbasnejad, I., Dean, D., & Fookes, C. (2018) · QUT ePrints (Queensland University of Technology) · Cited by 21
Dağaşan, Y., Juda, P., & Renard, P. (2020) · Ground Water · doi:10.1111/gwat.13005 · Cited by 21
“The subsequent training would then involve fine-tuning of the parameters in a shorter period of training time for the specific application (Ng et al. 2015).”
BackgroundFacial Expression Analysis and Its Potentials in IoT Systems: A Contemporary Survey
Shangguan, Z., Dong, Y., Guo, S., Leung, V. C. M., Deen, M. J., & Hu, X. (2025) · ACM Computing Surveys · doi:10.1145/3737456 · Cited by 22
“For example, Ng et al. [148] used ImageNet [34] dataset to obtain the pre-trained models that are fine-tuned by the other related MaE datasets.”
Sierra, S., Ramo, R., Padilla, M., & Cobo, A. (2025) · Environmental Monitoring and Assessment · doi:10.1007/s10661-025-13870-5 · Cited by 30
“…of these areas The proportion of training data with respect to the territory’s extension was deliberately small to evaluate the generalization capability of the classification algorithms and analyze how data augmentation techniques can improve their results (Fujisawa et al., 2019; Ng et al., 2015).”
Emotion Recognition in Conversations with Transfer Learning from Generative Conversation Modeling.
Hazarika, D., Poria, S., Zimmermann, R., & Mihalcea, R. F. (2019) · Cited by 20
Deep Learning-Based Emotion Recognition from Real-Time Videos
Zhou, W., Cheng, J., Lei, X., Beneš, B., & Adamo‐Villani, N. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-49062-1_22 · Cited by 21
“Several approaches have been introduced for robust head rotation detection [53], detection of facial features [64], speech [19], or even emotions [44].”
Backgroundda Silva, S., Yano, M. O., & Gonsalez, C. G. (2021) · Journal of Nondestructive Evaluation · doi:10.1007/s10921-021-00794-6 · Cited by 20
“The application of this methodology extends to different areas of research, such as image processing and classification [23,25], natural language processing [22,46], medical diagnosis [35], etc.”
MethodFacial Expression Recognition for the Blind Using Deep Learning
Joseph, J. L., & Mathew, S. P. (2021) · IEEE International Conference on Computing, Power and Communication Technologies (IC2PCT) · doi:10.1109/gucon50781.2021.9574035 · Cited by 20
“The transfer learning approach was based on the pre trained mobile net model [10][12].”
MethodCan deep learning assist automatic identification of layered pigments from XRF data?
Xu, B., Wu, Y., Hao, P., Vermeulen, M., McGeachy, A. C., Smith, K., Eremin, K., Rayner, G., Verri, G., Willomitzer, F., et al. (2022) · Journal of Analytical Atomic Spectrometry · doi:10.1039/d2ja00246a · Cited by 20
Machine Learning Techniques for Real-Time Emotion Detection from Facial Expressions
Sharma, A., Bajaj, V., & Arora, J. (2023) · IEEE Delhi Section Conference (DELCON) · doi:10.1109/delcon57910.2023.10127369 · Cited by 20
“The fully functional system captures the emotional state of the person in public places to avoid any illegal concerns [6].”
BackgroundPinitas, K., Renaudie, D., Thomsen, M., Barthet, M., Makantasis, K., Liapis, A., & Yannakakis, G. N. (2023) · INTERNATIONAL CONFERENCE ON MULTIMODAL INTERACTION · doi:10.1145/3577190.3614203 · Cited by 20
“[38] used CNNs pretrained on ImageNet [45] to perform emotion recognition on small datasets.”
MethodAttention-enhanced optimized deep ensemble network for effective facial emotion recognition
Khan, T., Yasir, M., & Choi, C. (2025) · Alexandria Engineering Journal · doi:10.1016/j.aej.2025.01.078 · Cited by 20
Ali, A. E., Wallbaum, T., Wasmann, M., Heuten, W., & Boll, S. (2017) · CHI Conference Extended Abstracts on Human Factors in Computing Systems (CHI EA) · doi:10.1145/3027063.3053086 · Cited by 20
“, using transferbased learning to deal with our small dataset [21]).”
MethodHuman Emotion Recognition in Video Using Subtraction Pre-Processing
He, Z., Tian, J. W., Basu, A., Soraghan, J. J., Caterina, G. D., & Petropoulakis, L. (2019) · International Conference on Machine Learning and Computing (ICMLC) · doi:10.1145/3318299.3318321 · Cited by 20
“In 2015 Emotion Recognition in the Wild (EmotiW) Challenge [7], the results in the paper show that RNNCNN system has better performance then deep learning CNN model.”
BackgroundEmotion Recognition in Gamers Wearing Head-mounted Display
Yong, H., Lee, J., & Choi, J. (2019) · IEEE Conference on Virtual Reality and 3D User Interfaces (VR) · doi:10.1109/vr.2019.8797736 · Cited by 20
“Feature points were selected or automatically extracted from the face and various techniques from nearest neighbor method to Convolutional Neural Networks (CNNs) were employed [1,8].”
BackgroundMethodA Deep Framework for Facial Emotion Recognition using Light Field Images
Sepas‐Moghaddam, A., Etemad, A., Correia, P. L., & Pereira, F. M. B. (2019) · International Conference on Affective Computing and Intelligent Interaction and workshops · doi:10.1109/acii.2019.8925445 · Cited by 20
“Nowadays, the most efficient and commonly used CNN architectures for facial emotion recognition are AlexNet [39], VGG-Very-Deep-16 [40], and VGG-Very-Deep19 [41].”
BackgroundExpression recognition with deep features extracted from holistic and part-based models
Happy, S. L., Dantcheva, A., & Brémond, F. (2020) · Image and Vision Computing · doi:10.1016/j.imavis.2020.104038 · Cited by 20
“Ng et al. [39] observed the improvement in expression classification by using the model weights trained on ImageNet.”
MethodMultimodal-multisensor affect detection
D’Mello, S. K., Bosch, N., & Chen, H. (2018) · ACM eBooks · doi:10.1145/3107990.3107998 · Cited by 20
“For example, Ng et al. [2015] found a 16% improvement by fine-tuning an object recognition deep network using multiple affect databases versus training on only one affect database.”
Shapiee, M. N. A., Ibrahim, M. A. R., Razman, M. A. M., Abdullah, M. A., Musa, R. M., & Majeed, A. P. P. A. (2020) · Lecture notes in electrical engineering · doi:10.1007/978-981-15-6025-5_20 · Cited by 20
Deep CNN-based Feature Extractor for Target Recognition in Thermal Images
Akula, A., & Sardana, H. K. (2019) · IEEE Region 10 Conference (TENCON) · doi:10.1109/tencon.2019.8929697 · Cited by 19
“Researchers have reported enhanced classification performance using models pre-trained on millions of images from the ImageNet database as a feature extractor for a number of applications such as emotion recognition [33], human action recognition [34], tumor classification in mammography images [35].”
BackgroundAdvances in facial expression recognition technologies for emotion analysis
Jayaswal, R., Ansari, M. A., Dixit, M., Singh, D. K., & Ahmad, S. (2025) · Discover Computing · doi:10.1007/s10791-025-09699-8 · Cited by 25
Rytky, S. J. O., Huang, L., Tanska, P., Tiulpin, A., Panfilov, E., Herzog, W., Korhonen, R. K., Saarakkala, S. S., & Finnilä, M. (2021) · Journal of Anatomy · doi:10.1111/joa.13435 · Cited by 20
Hybrid-Learning-Based Operational Visual Quality Inspection for Edge-Computing-Enabled IoT System
Chu, Y., Feng, D., Liu, Z., Zhao, Z., Wang, Z., Xia, X., & Quek, T. Q. S. (2021) · IEEE Internet of Things Journal · doi:10.1109/jiot.2021.3107902 · Cited by 20
Link, J., Perst, T., Stoeve, M., & Eskofier, B. M. (2022) · Sensors · doi:10.3390/s22072560 · Cited by 18
“In a second step, model parameters are fine-tuned on the target dataset [7].”
BackgroundVisual attention based composite dense neural network for facial expression recognition
Shaik, N. S., & Cherukuri, T. K. (2022) · Journal of Ambient Intelligence and Humanized Computing · doi:10.1007/s12652-022-03843-8 · Cited by 18
“Transfer Learning for FER systems achieved better performance (Ng et al. 2015).”
CSI-Based Human Activity Recognition Using Multi-Input Multi-Output Autoencoder and Fine-Tuning
Chahoushi, M., Nabati, M., Asvadi, R., & Ghorashi, S. A. (2023) · Sensors · doi:10.3390/s23073591 · Cited by 20
“Similar to audio-visual tasks [25], which aim to detect emotions, MIMO-AE could receive two inputs with different natures.”
Backgroundm_AutNet–A Framework for Personalized Multimodal Emotion Recognition in Autistic Children
Kurian, A., & Tripathi, S. (2024) · IEEE Access · doi:10.1109/access.2024.3403087 · Cited by 21
Facial Expression Classification using Deep Extreme Inception Networks
Raksarikorn, T., & Kangkachit, T. (2018) · International Joint Conference on Computer Science and Software Engineering · doi:10.1109/jcsse.2018.8457396 · Cited by 18
“The tranfer learning approach was applied to CNN architectures in [6].”
MethodData Driven Feature Extraction for Gender Classification using Multi-Script Handwritten Texts
Moetesum, M., Siddiqi, I., Djeddi, C., Hannad, Y., & Al-Maadeed, S. A. (2018) · International Conference on Frontiers in Handwriting Recognition · doi:10.1109/icfhr-2018.2018.00104 · Cited by 18
“While CNNs can be trained from scratch for a given problem, it is also common to employ pre-trained CNNs either as feature extractors [22] or fine tune them to a specific problem [37], [42] by continuing back propagation.”
BackgroundEmotion recognition and drowsiness detection using Python
Uppal, A., Tyagi, S., Kumar, R., & Sharma, S. (2019) · International Conference on Cloud Computing, Data Science & Engineering (Confluence) · doi:10.1109/confluence.2019.8776617 · Cited by 18
“emotions[10]”
Training Deep Neural Networks for Small and Highly Heterogeneous MRI Datasets for Cancer Grading
Wodziński, M., Banzato, T., Atzori, M., Andrearczyk, V., Cid, Y. D., & Müller, H. (2020) · Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) · doi:10.1109/embc44109.2020.9175634 · Cited by 18
“In [6], the authors combined transfer learning with domain adaptation, to perform emotion recognition using a small and heterogeneous data set.”
BackgroundA Comparative Study of Deep Learning Methods on Food Classification Problem
Memiş, S., Arslan, B., Batur, O. Z., & Sönmez, E. B. (2020) · Innovations in Intelligent Systems and Applications Conference (ASYU) · doi:10.1109/asyu50717.2020.9259904 · Cited by 18
Epileptic seizure detection using convolutional neural networks and recurrence plots of EEG signals
Ravi, S., Shreenidhi, S., Shahina, A., Ilakiyaselvan, N., & Khan, A. N. (2022) · Multimedia Tools and Applications · doi:10.1007/s11042-021-11608-2 · Cited by 18
“[18] demonstrated the effectiveness of transfer learning where weights from a pre-trained model can be used to fine-tune the network for the desired task and improve accuracy even when the dataset is small.”
BackgroundXu, Y. W., Liu, J., Zhai, Y., Gan, J., Zeng, J., Cao, H., Scotti, F., Piuri, V., & Labati, R. D. (2019) · Soft Computing · doi:10.1007/s00500-019-04530-1 · Cited by 18
Probabilistic Attribute Tree in Convolutional Neural Networks for Facial Expression RecognitionPreprint
Cai, J., Meng, Z., Khan, A.-S., Li, Z., O’Reilly, J. D., & Tong, Y. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1812.07067 · Cited by 18
Parekh, V., Foong, P. S., Zhao, S., & Subramanian, R. (2018) · ACM International Conference on Intelligent User Interfaces (IUI) · doi:10.1145/3172944.3173010 · Cited by 17
“The Emotion Detection module implements the deep network for emotion recognition described in [12].”
BackgroundMethodKilfoil, J. P., Rodriguez-Pinto, I. I., Kiszka, J. J., Heithaus, M. R., Zhang, Y., Roa, C., Ailloud, L., Campbell, M. D., & Wirsing, A. J. (2020) · ICES Journal of Marine Science · doi:10.1093/icesjms/fsaa161 · Cited by 17
Leveraging universality of jet taggers through transfer learning
Dreyer, F., Grabarczyk, R., & Monni, P. F. (2022) · The European Physical Journal C · doi:10.1140/epjc/s10052-022-10469-9 · Cited by 21
Supervised Contrastive Learning for Affect Modelling
Pinitas, K., Makantasis, K., Liapis, A., & Yannakakis, G. N. (2022) · INTERNATIONAL CONFERENCE ON MULTIMODAL INTERACTION · doi:10.1145/3536221.3556584 · Cited by 17
“[33] used a deep CNN pre-trained on the generic ImageNet dataset and leveraged transfer learning techniques to perform emotion recognition on small datasets.”
MethodMaliński, K., & Okarma, K. (2023) · Electronics · doi:10.3390/electronics12112449 · Cited by 24
Emotional Stability Detection Using Convolutional Neural Networks
Hussein, E. S. A., Qidwai, U. A., & Al-Meer, M. H. (2020) · IEEE International Conference on Informatics, IoT, and Enabling Technologies (ICIoT) · doi:10.1109/iciot48696.2020.9089440 · Cited by 17
“The proposed model in [16] uses AlexNet and CNN-M2048 model for FER; they propose a cascaded-finetuning process which shows better results compared to finetuning over a cross-dataset.”
MethodActive AU Based Patch Weighting for Facial Expression Recognition
Xie, W., Shen, L., Yang, M., & Lai, Z. (2017) · Sensors · doi:10.3390/s17020275 · Cited by 15
“Meanwhile, CNN-based algorithms require a large number of training parameters and high computational complexity [16,17].”
InfluentialAlzamzami, F., & Saddik, A. E. (2023) · IEEE Access · doi:10.1109/access.2023.3274744 · Cited by 17
Sharma, C., Sharma, S., Sakshi, S., & Chen, Y.-H. (2024) · Discover Applied Sciences · doi:10.1007/s42452-024-06217-1 · Cited by 16
Further Understanding Videos through Adverbs: A New Video Task
Pang, B., Zha, K., Zhang, Y., & Lu, C. (2020) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v34i07.6855 · Cited by 17
“For action recognition tasks, some works extract features from video frames separately and then fuse them together, where RNN is widely used (Donahue et al. 2015; Wu et al. 2015; Srivastava, Mansimov, and Salakhudinov 2015; Pang et al. 2019) and many pooling methods have been developed (Yue-Hei Ng et al. 2015; Karpath…”
BackgroundMethodInfluentialDeep Transfer Learning For Plant Center Localization
Cai, E., Baireddy, S., Yang, C., Crawford, M. M., & Delp, E. J. (2020) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw50498.2020.00039 · Cited by 17
“ature extractor for the input image [18]. For example, in [19], Oquab et al. transfer the weight of a pretrained CNN to improve the performance of the network with a small amount of training data. In [20], Ng et al. fine-tune a pretrained CNN for emotion recognition on small datasets. Tapas et al. [21] retrain a pretra…”
A web crowdsourcing framework for transfer learning and personalized Speech Emotion Recognition
Vryzas, N., Vrysis, L., Kotsakis, R., & Dimoulas, C. A. (2021) · Machine Learning with Applications · doi:10.1016/j.mlwa.2021.100132 · Cited by 16
Light Deep Models for Cognitive Computing in Intelligent Transportation Systems
Muhammad, G., & Hossain, M. S. (2022) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2022.3171913 · Cited by 16
“The authors in [11] used a transfer learning CNN model trained on an image classification task to realize facial emotions.”
MethodUsing Emotion Recognition in Intelligent Interface Design for Elderly Care
Khanal, S. R., Reis, A., Barroso, J., & Filipe, V. M. D. J. (2018) · Advances in intelligent systems and computing · doi:10.1007/978-3-319-77712-2_23 · Cited by 16
Deep Residual Learning for Analyzing Customer Satisfaction using Video Surveillance
Sugianto, N., Tjondronegoro, D., & Tydd, B. (2018) · IEEE International Conference on Advanced Video and Signal Based Surveillance (AVSS) · doi:10.1109/avss.2018.8639478 · Cited by 16
“Previous studies, such as [13] and [22], confirmed that knowledge transfer and augmenting data from different domain datasets approaches can be used to build deeplearning models with a small dataset.”
BackgroundFacial Expression Recognition by Transfer Learning for Small Datasets
Li, J., Huang, S., Zhang, X., Fu, X., Chang, C., Tang, Z., & Luo, Z. (2019) · Advances in intelligent systems and computing · doi:10.1007/978-3-030-16946-6_62 · Cited by 16
Video-based emotion recognition in the wild
Salah, A. A., Kaya, H., & Gürpınar, F. (2018) · Elsevier eBooks · doi:10.1016/b978-0-12-814601-9.00031-6 · Cited by 15
Analyzing Group-Level Emotion with Global Alignment Kernel based Approach
Huang, X., Dhall, A., Goecke, R., Pietikäinen, M., & Zhao, G. (2019) · IEEE Transactions on Affective Computing · doi:10.1109/taffc.2019.2953664 · Cited by 15
“and poor illumination, we used deep convolutional neural network feature [45], [46] as the high-level feature, and Rieszbased Volume Local Binary Pattern (RVLBP) [18] as the lowlevel feature for global alignment kernel.”
MethodFacial expression recognition based on the ensemble learning of CNNs
Chen, J., Li, C. L., & Ying, Z. (2020) · IEEE International Conference on Signal Processing, Communications and Computing (ICSPCC) · doi:10.1109/icspcc50002.2020.9259543 · Cited by 17
“Hong-Wei Ng [6] and others adopted the method of transfer learning, The final recognition rate is 55.6%.”
MethodInfluentialFusion of Machine Learning and Privacy Preserving for Secure Facial Expression Recognition
Ullah, A., Wang, J., Anwar, M. S., Ahmad, A., Nazir, S., Khan, H. U., & Fei, Z. (2021) · Security and Communication Networks · doi:10.1155/2021/6673992 · Cited by 15
Suguitan, M., Gómez, R., & Hoffman, G. (2020) · ACM/IEEE International Conference on Human-Robot Interaction (HRI) · doi:10.1145/3319502.3374807 · Cited by 21
“Neural networks have found success in complex applications for affective computing, primarily in perceptual tasks such as emotion recognition [12, 20], though some works have explored using neural networks for affective speech and expression generation [5, 33].”
BackgroundWikanningrum, A., Rachmadi, R. F., & Ogata, K. (2019) · 2019 International Conference on Computer Engineering, Network, and Intelligent Multimedia (CENIM) · doi:10.1109/cenim48368.2019.8973312 · Cited by 17
“[14] perform transfer learning from ImageNet weights for emotion recognition.”
MethodEncrypted optical information in nonlinear chaotic systems uncovered using neural networks
Konstantakis, P., Manousidaki, M., & Tzortzakis, S. (2024) · Optica · doi:10.1364/optica.530643 · Cited by 14
DeepDA-Ace: A Novel Domain Adaptation Method for Species-Specific Acetylation Site Prediction
Liu, Y., Wang, Q., & Xi, J. (2022) · Mathematics · doi:10.3390/math10142364 · Cited by 14
A Comparative Study of SVM, CNN, and DCNN Algorithms for Emotion Recognition and Detection
Prabha, R., A, S. G., Razmah, M., Akshaya, S. R., Sivashree, J., & Swathi, J. C. (2022) · Algorithms for intelligent systems · doi:10.1007/978-981-19-6004-8_64 · Cited by 14
From the Lab to the Wild: Affect Modeling Via Privileged Information
Makantasis, K., Pinitas, K., Liapis, A., & Yannakakis, G. N. (2023) · IEEE Transactions on Affective Computing · doi:10.1109/taffc.2023.3265072 · Cited by 14
“Jung et al. [31] boosted the performance of facial-based affect models by exploiting high-level spatiotemporal representations of facial action points produced by CNNs. Ng et al. [32] proposed transfer of learning across CNNs for emotion recognition through visual cues.”
BackgroundShah, Z. H., Muller, M. W., Hübner, W., Wang, T.-C., Telman, D., Huser, T. R., & Schenck, W. (2024) · GigaScience · doi:10.1093/gigascience/giad109 · Cited by 14
“In addition, transfer learning is also very helpful when it comes to alleviating large dataset r equir ements [ 50 ].”
Improved facial expression recognition method based on ROI deep convolutional neutral network
Sun, X. N., Lv, M., Quan, C., & Ren, F. (2017) · International Conference on Affective Computing and Intelligent Interaction and workshops · doi:10.1109/acii.2017.8273609 · Cited by 14
“Face images preprocessing proves to be a crucial step in order to acquire good recognition performance [6].”
BackgroundBarros, P. V. A., Barakova, E. I., & Wermter, S. (2020) · IEEE Transactions on Affective Computing · doi:10.1109/taffc.2020.3002657 · Cited by 14
Thermal Face Authentication with Convolutional Neural Network
Sayed, M., & Baker, F. (2018) · Journal of Computer Science · doi:10.3844/jcssp.2018.1627.1637 · Cited by 14
“The proposed method is known in the literature as a transfer learning approach and has been successful in many applications (Hoo-Chang et al ., 2016), (Ng et al ., 2015), (Wang and Deng, 2018) and (Tajbakhsh et al ., 2017).”
BackgroundMethodTransfer Learning From Sound Representations For Anger Detection in SpeechPreprint
Elshaer, M. E. A., Wisdom, S., & Mishra, T. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1902.02120 · Cited by 14
“In the vision modality, Ng et al. [3] used a deep convolutional network initialized with a model pre-trained on ImageNet [4], then fine-tuned on datasets relevant to facial expressions, followed by fine-tuning on the EmotiW 2015 dataset [5].”
MethodA Computer Vision-Based Method for Classification of Red Meat Quality After Nitrosamine Appendage
Arora, M., & Mangipudi, P. (2021) · International Journal of Computational Intelligence and Applications · doi:10.1142/s146902682150005x · Cited by 13
Artificial Intelligence in Higher Education
Jupalli, T. K., Reddy, M. S. T., & Kondaveeti, H. K. (2022) · Advances in educational technologies and instructional design book series · doi:10.4018/978-1-6684-5400-8.ch001 · Cited by 16
Steel Plate Fault Detection Using the Fitness-Dependent Optimizer and Neural Networks
Farahmand‐Tabar, S., & Rashid, T. A. (2023) · Springer eBooks · doi:10.1007/978-981-19-8851-6_41-1 · Cited by 13
Aly, S. F., & Abbott, A. L. (2019) · Lecture notes in computer science · doi:10.1007/978-3-030-20205-7_9 · Cited by 13
Classification Method of Plug Seedlings Based on Transfer Learning
Zhang, X., Tan, Y., Liu, X., & Yang, S. (2019) · Applied Sciences · doi:10.3390/app9132725 · Cited by 13
“Convolutional neural network is a robust method for classification [7–9], which has been widely verified in the fields of agricultural product classification and agricultural disease identification [10–13].”
MethodExpression Classification in Children Using Mean Supervised Deep Boltzmann Machine
Nagpal, S., Singh, M., Mayank Vatsa, A., Singh, R., & Afzel Noore, B. (2019) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw.2019.00033 · Cited by 13
“Ng et al. [30] presented a transfer learning based approach, where CNNs pre-trained on the ImageNet dataset [9] undergo a two-step supervised fine-tuning process.”
MethodAmbroa, E. M., Perez-Alija, J., & Gallego, P. (2021) · Medical dosimetry · doi:10.1016/j.meddos.2021.03.005 · Cited by 12
A systematic review on handwritten document analysis and recognition
Inunganbi, S. C. (2023) · Multimedia Tools and Applications · doi:10.1007/s11042-023-15326-9 · Cited by 14
“In [79], TL (AlexNet and VggNet) is employed for identification of human emotions.”
Recognition of printed Urdu ligatures using convolutional neural networks
Uddin, I., Javed, N., Siddiqi, I., Khalid, S., & Khurshid, K. (2019) · Journal of Electronic Imaging · doi:10.1117/1.jei.28.3.033004 · Cited by 12
Tang, C., Uriarte, M., Jin, H., Morton, D. C., & Zheng, T. (2021) · Methods in Ecology and Evolution · doi:10.1111/2041-210x.13549 · Cited by 13
“…such as AlexNet (Krizhevsky et al., 2012), VGG network (Simonyan & Zisserman, 2014) and ResNet (He et al., 2016) are well-suited for learning expressive image representations, as they have achieved excellent generalization performance in transfer learning (Ng et al., 2015; Shin et al., 2016).”
BackgroundGittler, T., Relea, E., Corti, D., Giorgio Corani, A., Weiss, L., Cannizzaro, D., & Wegener, K. (2019) · Procedia CIRP · doi:10.1016/j.procir.2019.02.026 · Cited by 13
Facial Expression Recognition Using Disentangled Adversarial LearningPreprint
Ali, K., & Hughes, C. E. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1909.13135 · Cited by 13
A transfer learning-based deep learning approach for automated COVID-19 diagnosis with audio data
Akgün, D., Kabakuş, A. T., Şentürk, Z. K., Şentürk, A., & Küçükkülahlı, E. (2021) · TURKISH JOURNAL OF ELECTRICAL ENGINEERING & COMPUTER SCIENCES · doi:10.3906/elk-2105-64 · Cited by 12
Gemstone classification using ConvNet with transfer learning and fine-tuning
Freire, W. M., Amaral, A. M. M. M., & Costa, Y. M. G. (2022) · International Conference on Systems, Signals and Image Processing · doi:10.1109/iwssip55020.2022.9854441 · Cited by 12
de Santana, M. A., Fonseca, F. S., Torcate, A. S., & dos Santos, W. P. (2023) · Research on Biomedical Engineering · doi:10.1007/s42600-023-00293-9 · Cited by 15
“The FER database was also used by (Ng et al. 2015).”
MethodLocal Subclass Constraint for Facial Expression Recognition in the Wild
Luo, Z., Hu, J., & Deng, W. (2018) · Proceedings - International Conference on Pattern Recognition/Proceedings/International Conference on Pattern Recognition · doi:10.1109/icpr.2018.8545847 · Cited by 12
“Benefiting from the development of deep learning, especially the success of DCNNs in computer vision field, robust expression features can be learned from labeled training data and DCNNs have achieved promising performance in real world facial expression recognition[18], [19].”
Background
Metadata
- Authors
- Ng, H.-W., Nguyen, V. D., Vonikakis, V., & Winkler, S.
- Affiliations
- University of Illinois at Urbana-Champaign (Singapore), Singapore, Singapore
- University of Illinois Urbana-Champaign (US)
- Venue
- Proceedings of the 2015 ACM on International Conference on Multimodal Interaction
- Publisher
- ACM
- Published
- 2015-11-09
- Pages
- 443-449
- OpenAlex
- W2253728219
- Cited by
- 657
- References
- 32
- Citations by year
- 2026: 13 · 2025: 29 · 2024: 37 · 2023: 70 · 2022: 69 · 2021: 106 · 2020: 106 · 2019: 106 · 2018: 63 · 2017: 44 · 2016: 13 · 2015: 1
- FWCI
- 53.84
- Type
- conference paper
- Language
- en
- Open access
- closed
- Locations
- INTERNATIONAL CONFERENCE ON MULTIMODAL INTERACTION (publishedVersion)
- Funders
- Agency for Science Technology and Research
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Emotion and Mood Recognition
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Face recognition and analysis
- Physical Sciences › Computer Science › Artificial Intelligence › Domain Adaptation and Few-Shot Learning
- Keywords
- emotion recognition, facial expression recognition, still images, transfer learning, convolutional neural networks, fine-tuning, ImageNet pre-training, small datasets
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
