- Improving Knowledge Distillation via Category Structure
- Describing Textures using Natural Language
- Learning and Aggregating Deep Local Descriptors for Instance-level Recognition
- Spatially Adaptive Inference with Stochastic Feature Sampling and Interpolation
- Gradient Centralization: A New Optimization Technique for Deep Neural Networks
- Diffraction Line Imaging
- Self-Challenging Improves Cross-Domain Generalization
- Multitask Learning Strengthens Adversarial Robustness
- Volumetric Transformer Networks
- Patch-wise Attack for Fooling Deep Neural Network
- Supervised Edge Attention Network for Accurate Image Instance Segmentation
- Layer-wise Conditioning Analysis in Exploring the Learning Dynamics of DNNs
- Efficient Scale-Permuted Backbone with Learned Resource Distribution
- Rethinking Bottleneck Structure for Efficient Mobile Network Design