- Cascade R-CNN: Delving into High Quality Object Detection
- Batch Size in Deep Learning
- DeepLab V3+: Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation
- Label Refinery: Improving ImageNet Classification through Label Progression
- Uncertainty and Deep Learning
- Progress & Compress: A scalable framework for continual learning
- Batch-Instance Normalization
- Geometry-aware CNN Features
- Discover New Value of Medical Imaging using Deep Learning
- DSD : Dense-Sparse-Dense Training for Deep Neural Networks
