During the training process, deep neural networks implicitly learn to represent the input data samples through a hierarchy of features, where the size of the hierarchy is determined by the number of layers. In this paper, we focus on enforcing the discriminative power of the high-level representations, that are typically learned by the deeper layers (closer to the output). To this end, we introduce a new loss term inspired by the Gini impurity, which is aimed at minimizing the entropy (increasing the discriminative power) of individual high-level features with respect to the class labels. Although our Gini loss induces highly-discriminative features, it does not ensure that the distribution of the high-level features matches the distribution of the classes. As such, we introduce another loss term to minimize the Kullback-Leibler divergence between the two distributions. We conduct experiments on two image classification data sets (CIFAR-100 and Caltech 101), considering multiple neural architectures ranging from convolutional networks (ResNet-17, ResNet-18, ResNet-50) to transformers (CvT). Our empirical results show that integrating our novel loss terms into the training objective consistently outperforms the models trained with cross-entropy alone, without increasing the inference time at all.
Auto-weighted low-rank representation for clustering
Institute of Information Science, Beijing Jiaotong University（北京交通大学信息科学研究所）；Beijing Key Laboratory of Advanced Information Science and Network Technology（北京先进信息科学与网络技术重点实验室）；Department of Computer Science and Technology, Tsinghua University（清华大学计算机科学与技术系）
DDG-Net: Discriminability-Driven Graph Network for Weakly-supervised Temporal Action Localization
Beijing University of Posts and Telecommunications（北京邮电大学）；Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所）；Centre for Artificial Intelligence and Robotics, HKISI_CAS（中国科学院香港创新研究院人工智能与机器人创新中心）；University of Chinese Academy of Sciences（中国科学院大学）
COAL: Counterfactual and Observation-Enhanced Alignment Learning for Discriminative Referring Multi-Object Tracking
School of Automation, Southeast University, Nanjing, China（东南大学自动化学院，南京，中国）；Key Laboratory of Measurement and Control of Complex Systems of Engineering, Ministry of Education, Nanjing, China（工程复杂系统测量与控制国家重点实验室，教育部，南京，中国）；School of Physical & Mathematical Sciences, Nanyang Technological University, Singapore（南洋理工大学物理与数学科学学院，新加坡）；Big Data Institute, Central South University, Changsha, China（中南大学大数据研究院，长沙，中国）
SuSana Distancia is all you need: Enforcing class separability in metric learning via two novel distance-based loss functions for few-shot image classification
Tecnologico de Monterrey, School of Engineering and Sciences（蒙特雷理工学院工程与科学学院）；Centro de Investigación y de Estudios Avanzados（高级研究与调查中心）
PC-UNet: An Enforcing Poisson Statistics U-Net for Positron Emission Tomography Denoising
School of Computer Science and Technology, Guangdong University of Technology（广东技术大学计算机科学与技术学院）；Zhongshan School of Medicine, Sun Yat-sen University（中山大学仲沙医学院）；School of Mathematics and Statistics, Guangdong University of Technology（广东技术大学数学与统计学院）；School of Mathematics (Zhuhai), Sun Yat-sen University（中山大学珠海数学学院）；Future Technology Institute, South China University of Technology（华南理工大学未来技术研究院）；School of Computer Science and Engineering, Huizhou University（惠州大学计算机科学与工程学院）
Ensemble Soft-Margin Softmax Loss for Image Classification
Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所）；University of Chinese Academy of Sciences（中国科学院大学）；Beihang University（北京航空航天大学）；Tianjin University（天津大学）；Macau University of Science and Technology（澳门科技大学）
Gradually Vanishing Gap in Prototypical Network for Unsupervised Domain Adaptation
Anhui University（安徽大学）；University of Science and Technology of China（中国科学技术大学）；Chongqing University of Technology（重庆理工大学）；Hefei University of Technology（合肥工业大学）；Hebei University of Technology（河北工业大学）
Knowledge distillation via adaptive instance normalization
Computer Vision Laboratory, University of Nottingham, UK（诺丁汉大学计算机视觉实验室）；Samsung AI Center, Cambridge, UK（三星人工智能中心剑桥）；Queen Mary University of London, London, UK（伦敦玛丽女王大学）
NeurCADRecon: Neural Representation for Reconstructing CAD Surfaces by Enforcing Zero Gaussian Curvature
Shandong University（山东大学）；The University of Hong Kong（香港大学）；Qingdao University of Science and Technology（青岛科技大学）；AMSS, Chinese Academy Of Sciences（中国科学院数学与系统科学研究院）；Texas A&M University（德克萨斯农工大学）
D2-Net: Weakly-Supervised Action Localization via Discriminative Embeddings and Denoised Activations
Inception Institute of Artificial Intelligence（启元人工智能研究院）；Mohamed Bin Zayed University of AI（穆罕默德·本·扎耶德人工智能大学）；Monash University（莫纳什大学）；Linköping University（林雪平大学）；University of California Merced（加州大学默塞德分校）；Google Research（谷歌研究院）；Yonsei University（延世大学）
