{"type":"ScholarlyArticle","title":"Better Embedding and More Shots for Few-shot Learning","authors":[{"name":"Ziqiu Chi","slug":"ziqiu-chi","given":"Ziqiu","family":"Chi","affiliation":"Key Laboratory of Smart Manufacturing in Energy Chemical Process , East China University of Science and Technology , China; School of Information Science and Engineering , East China University of Science and Technology , China"},{"name":"Zhe Wang","slug":"zhe-wang","given":"Zhe","family":"Wang","affiliation":"Key Laboratory of Smart Manufacturing in Energy Chemical Process , East China University of Science and Technology , China; School of Information Science and Engineering , East China University of Science and Technology , China"},{"name":"Mengping Yang","slug":"mengping-yang","given":"Mengping","family":"Yang","affiliation":"Key Laboratory of Smart Manufacturing in Energy Chemical Process , East China University of Science and Technology , China; School of Information Science and Engineering , East China University of Science and Technology , China"},{"name":"Wei Guo","slug":"wei-guo","given":"Wei","family":"Guo","affiliation":"Key Laboratory of Smart Manufacturing in Energy Chemical Process , East China University of Science and Technology , China; School of Information Science and Engineering , East China University of Science and Technology , China"},{"name":"Xinlei Xu","slug":"xinlei-xu","given":"Xinlei","family":"Xu","affiliation":"Key Laboratory of Smart Manufacturing in Energy Chemical Process , East China University of Science and Technology , China; School of Information Science and Engineering , East China University of Science and Technology , China"}],"abstract":"In few-shot learning, methods are enslaved to the scarce labeled data, resulting in suboptimal embedding. Recent studies learn the embedding network by other large-scale labeled data. However, the trained network may give rise to the distorted embedding of target data. We argue two respects are required for an unprecedented and promising solution. We call them Better Embedding and More Shots (BEMS). Suppose we propose to extract embedding from the embedding network. BE maximizes the extraction of general representation and prevents over-fitting information. For this purpose, we introduce the topological relation for global reconstruction, avoiding excessive memorizing. MS maximizes the relevance between the reconstructed embedding and the target class space. In this respect, increasing the number of shots is a pivotal but intractable strategy. As a creative method, we derive the bound of information-theory-based loss function and implicitly achieve infinite shots with negligible cost. A substantial experimental analysis is carried out to demonstrate the state-of-the-art performance. Compared to the baseline, our method improves by up to 10%+. We also prove that BEMS is suitable for both standard pre-trained and meta-learning embedded networks.","doi":"10.24963/ijcai.2022/398","url":"https://ijcai.org/proceedings/2022/398","pdf":"https://ijcai.org/proceedings/2022/0398.pdf","markdown":"https://ijcai.org/proceedings/2022/398.md","bibtex":"https://ijcai.org/proceedings/2022/bibtex/398","volume":{"slug":"2022","title":"Proceedings of the Thirty-First International Joint Conference on Artificial Intelligence","short_title":"IJCAI 2022","url":"https://ijcai.org/proceedings/2022/","year":2022,"location":"Vienna","dates":"23-29 July 2022","editors":"Luc De Raedt","isbn":"978-1-956792-00-3","publisher":"International Joint Conferences on Artificial Intelligence Organization","doi_prefix":"10.24963/ijcai.2022"},"section":"Machine Learning","track":"Main Track","number":398,"pages":{"first":2874,"last":2880},"keywords":["Machine Learning: Few-shot learning","Machine Learning: Theory of Deep Learning","Computer Vision: Machine Learning for Vision","Computer Vision: Transfer, low-shot, semi- and un- supervised learning","Machine Learning: Classification"],"videos":[{"label":"Video #1","url":"https://slideslive.com/38986449","length_sec":83,"provider":"slideslive"},{"label":"Video #2","url":"https://slideslive.com/38984592","length_sec":263,"provider":"slideslive"}],"references":[{"position":1,"raw":"[Cao et al., 2019] Tianshi Cao, Marc T Law, and Sanja Fi- dler. A theoretical analysis of the number of shots in few- shot learning. In ICLR, 2019.","title":"A theoretical analysis of the number of shots in fewshot learning","authors":"Cao","year":"2019","venue":"ICLR","doi":null,"url":null,"internal_url":null},{"position":2,"raw":"[Chen et al., 2019a] Wei-Yu Chen, Yen-Cheng Liu, Zsolt Ki- ra, Yu-Chiang Wang, and Jia-Bin Huang. A closer look at few-shot classification. In ICLR, 2019.","title":"A closer look at few-shot classification","authors":"Chen","year":"2019","venue":"ICLR","doi":null,"url":null,"internal_url":null},{"position":3,"raw":"[Chen et al., 2019b] Zitian Chen, Yanwei Fu, Yinda Zhang, Yu-Gang Jiang, Xiangyang Xue, and Leonid Sigal. Multi- level semantic feature augmentation for one-shot learning. TIP, 28(9):4594-4605, 2019.","title":"Multilevel semantic feature augmentation for one-shot learning","authors":"Chen","year":"2019","venue":"TIP","doi":null,"url":null,"internal_url":null},{"position":4,"raw":"[Chen et al., 2021] Da Chen, Yuefeng Chen, Yuhong Li, Feng Mao, Yuan He, and Hui Xue. Self-supervised learn- ing for few-shot image classification. In ICASSP, pages 1745-1749, 2021.","title":"Self-supervised learning for few-shot image classification","authors":"Yuefeng Da Chen, Yuhong Chen, Feng Li, Yuan Mao, Hui He, Xue","year":"2021","venue":"ICASSP","doi":null,"url":null,"internal_url":null},{"position":5,"raw":"[Chi et al., 2021] Ziqiu Chi, Zhe Wang, Mengping Yang, Dongdong Li, and Wenli Du. Learning to capture the query distribution for few-shot learning. TCSVT, 2021.","title":"Learning to capture the query distribution for few-shot learning","authors":"Chi","year":"2021","venue":"TCSVT","doi":null,"url":null,"internal_url":null},{"position":6,"raw":"[Finn et al., 2017] Chelsea Finn, Pieter Abbeel, and Sergey Levine. Model-agnostic meta-learning for fast adaptation of deep networks. In ICML, pages 1126-1135, 2017.","title":"Model-agnostic meta-learning for fast adaptation of deep networks","authors":"Finn","year":"2017","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":7,"raw":"[Guo and Cheung, 2020] Yiluan Guo and Ngai-Man Che- ung. Attentive weights generation for few shot learning via information maximization. In CVPR, pages 13499-13508, 2020.","title":"Yiluan Guo and Ngai-Man Cheung. Attentive weights generation for few shot learning via information maximization","authors":"Cheung Guo","year":"2020","venue":"CVPR","doi":null,"url":null,"internal_url":null},{"position":8,"raw":"[Hou and Sato, 2021] Mingcheng Hou and Issei Sato. A closer look at prototype classifier for few-shot image clas- sification. arXiv preprint arXiv:2110.05076, 2021.","title":"A closer look at prototype classifier for few-shot image classification","authors":"Sato Hou, Mingcheng Hou, Issei Sato","year":"2021","venue":null,"doi":null,"url":null,"internal_url":null},{"position":9,"raw":"[Lee and Chung, 2021] Dong Hoon Lee and Sae-Young Chung. Unsupervised embedding adaptation via early- stage feature reconstruction for few-shot classification. In ICML, pages 6098-6108, 2021.","title":"Unsupervised embedding adaptation via earlystage feature reconstruction for few-shot classification","authors":"Lee, Chung ; ] Dong, Hoon Lee, Sae-Young Chung","year":"2021","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":10,"raw":"[Li et al., 2020] Kai Li, Yulun Zhang, Kunpeng Li, and Yun Fu. Adversarial feature hallucination networks for few- shot learning. In CVPR, pages 13470-13479, 2020.","title":"Adversarial feature hallucination networks for fewshot learning","authors":"Kai Li, Yulun Zhang, Kunpeng Li, Yun Fu","year":"2020","venue":"CVPR","doi":null,"url":null,"internal_url":null},{"position":11,"raw":"[Liu et al., 2019] Yanbin Liu, Juho Lee, Minseop Park, Sae- hoon Kim, Eunho Yang, Sung Ju Hwang, and Yi Yang. Learning to propagate labels: Transductive propagation network for few-shot learning. In ICLR, 2019.","title":"Learning to propagate labels: Transductive propagation network for few-shot learning","authors":"Yanbin Liu, Juho Lee, Minseop Park, Saehoon Kim, Eunho Yang, Sung Ju Hwang, Yi Yang","year":"2019","venue":"ICLR","doi":null,"url":null,"internal_url":null},{"position":12,"raw":"[Liu et al., 2020] Jinlu Liu, Liang Song, and Yongqiang Qin. Prototype rectification for few-shot learning. In ECCV, pages 741-756, 2020.","title":"Prototype rectification for few-shot learning","authors":"Liu","year":"2020","venue":"ECCV","doi":null,"url":null,"internal_url":null},{"position":13,"raw":"[Mangla et al., 2020] Puneet Mangla, Nupur Kumari, Ab- hishek Sinha, Mayank Singh, Balaji Krishnamurthy, and Vineeth N Balasubramanian. Charting the right manifold: Manifold mixup for few-shot learning. In WACV, pages 2218-2227, 2020.","title":"Charting the right manifold: Manifold mixup for few-shot learning","authors":"Puneet Mangla, Nupur Kumari, Abhishek Sinha, Mayank Singh, Balaji Krishnamurthy, Vineeth N Balasubramanian","year":"2020","venue":"WACV","doi":null,"url":null,"internal_url":null},{"position":14,"raw":"[Park et al., 2020] Seong-Jin Park, Seungju Han, Ji-won Baek, Insoo Kim, Juhwan Song, Hae Beom Lee, Jae-Joon Han, and Sung Ju Hwang. Meta variance transfer: Learn- ing to augment from the others. In ICML, pages 7510- 7520, 2020.","title":"Meta variance transfer: Learning to augment from the others","authors":"Seong-Jin Park, Seungju Han, Ji-Won Baek, Insoo Kim, Juhwan Song, Hae Beom Lee, Jae-Joon Han, Sung Ju Hwang","year":"2020","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":15,"raw":"[Ren et al., 2018] Mengye Ren, Eleni Triantafillou, Sachin Ravi, Jake Snell, Kevin Swersky, Joshua B Tenenbaum, Hugo Larochelle, and Richard S Zemel. Meta-learning for semi-supervised few-shot classification. arXiv preprint arXiv:1803.00676, 2018.","title":"Meta-learning for semi-supervised few-shot classification","authors":"Ren","year":"2018","venue":null,"doi":null,"url":null,"internal_url":null},{"position":16,"raw":"[Snell et al., 2017] Jake Snell, Kevin Swersky, and Richard Zemel. Prototypical networks for few-shot learning. NeurIPS, 30, 2017.","title":"Prototypical networks for few-shot learning","authors":"Jake Snell, Kevin Swersky, Richard Zemel","year":"2017","venue":"NeurIPS","doi":null,"url":null,"internal_url":null},{"position":17,"raw":"[Tian et al., 2020] Yonglong Tian, Yue Wang, Dilip Krish- nan, Joshua B Tenenbaum, and Phillip Isola. Rethinking few-shot image classification: a good embedding is all you need? In ECCV, pages 266-282, 2020.","title":"Rethinking few-shot image classification: a good embedding is all you need?","authors":"Yonglong Tian, Yue Wang, Dilip Krishnan, Joshua B Tenenbaum, Phillip Isola","year":"2020","venue":"ECCV","doi":null,"url":null,"internal_url":null},{"position":18,"raw":"[Vinyals et al., 2016] Oriol Vinyals, Charles Blundell, Tim- othy Lillicrap, koray kavukcuoglu, and Daan Wierstra. Matching networks for one shot learning. In NeurIPS, vol- ume 29, 2016.","title":"Matching networks for one shot learning","authors":"Oriol Vinyals, Charles Blundell, Timothy Lillicrap, Daan Wierstra","year":"2016","venue":"NeurIPS","doi":null,"url":null,"internal_url":null},{"position":19,"raw":"[Wah et al., 2011] Catherine Wah, Steve Branson, Peter Welinder, Pietro Perona, and Serge Belongie. The caltech- ucsd birds-200-2011 dataset. 2011.","title":"The caltechucsd birds-200-2011 dataset","authors":"Catherine Wah, Steve Branson, Peter Welinder, Pietro Perona, Serge Belongie","year":"2011","venue":null,"doi":null,"url":null,"internal_url":null},{"position":20,"raw":"[Wang et al., 2019] Yan Wang, Wei-Lun Chao, Kilian Q Weinberger, and Laurens van der Maaten. Simpleshot: Re- visiting nearest-neighbor classification for few-shot learn- ing. arXiv preprint arXiv:1911.04623, 2019.","title":"Simpleshot: Revisiting nearest-neighbor classification for few-shot learning","authors":"Yan Wang, Wei-Lun Chao, Kilian Q Weinberger, Laurens Van Der Maaten","year":"2019","venue":null,"doi":null,"url":null,"internal_url":null},{"position":21,"raw":"[Wang et al., 2021] Yulin Wang, Gao Huang, Shiji Song, X- uran Pan, Yitong Xia, and Cheng Wu. Regularizing deep networks with semantic data augmentation. TPAMI, 2021.","title":"Regularizing deep networks with semantic data augmentation","authors":"Yulin Wang, Gao Huang, Shiji Song, Xuran Pan, Yitong Xia, Cheng Wu","year":"2021","venue":"TPAMI","doi":null,"url":null,"internal_url":null},{"position":22,"raw":"[Yang et al., 2020] Shuo Yang, Lu Liu, and Min Xu. Free lunch for few-shot learning: Distribution calibration. In ICLR, 2020.","title":"Free lunch for few-shot learning: Distribution calibration","authors":"Shuo Yang, Lu Liu, Min Xu","year":"2020","venue":"ICLR","doi":null,"url":null,"internal_url":null},{"position":23,"raw":"[Zhang et al., 2018] Ruixiang Zhang, Tong Che, Zoubin Ghahramani, Yoshua Bengio, and Yangqiu Song. Meta- gan: An adversarial approach to few-shot learning. In NeurIPS, volume 31, 2018.","title":"Metagan: An adversarial approach to few-shot learning","authors":"Zhang","year":"2018","venue":"NeurIPS","doi":null,"url":null,"internal_url":null},{"position":24,"raw":"[Ziko et al., 2020] Imtiaz Ziko, Jose Dolz, Eric Granger, and Ismail Ben Ayed. Laplacian regularized few-shot learning. In ICML, pages 11660-11670, 2020.","title":"Laplacian regularized few-shot learning","authors":"Imtiaz Ziko, Jose Dolz, Eric Granger, Ismail Ben, Ayed","year":"2020","venue":"ICML","doi":null,"url":null,"internal_url":null}],"license_note":"Copyright International Joint Conferences on Artificial Intelligence."}