I wrote the paper then Triet helped me with the revisions. A survey of deep language models and their applications in software engineering. Paper got accepted by Computing Survey.

Contributed equally with the second author. In this work, we achieve improved mask prediction by effectively combining instance-level information with semantic information with lower-level fine-granularity. Our main contribution is a blender module which draws inspiration from both top-down and bottom-up instance segmentation approaches. The proposed BlendMask can effectively predict dense per-pixel position-sensitive instance features with very few channels, and learn attention maps for each instance with merely one convolution layer, thus being fast in inference. BlendMask can be easily incorporated with the state-of-the-art one-stage detection frameworks and outperforms Mask R-CNN under the same training schedule while being 20% faster.

Haokui adapted my implementation of AutoDeepLab for low-level image processing tasks. One highlight of the framework is that it save two thirds of the memory required for AutoDeepLab.

Contributed equally with the first two authors. We built upon our CVPR19 work for object detection on COCO. At first I thought it would be easy but we met many obstacles mostly about training efficiency. Even up to now I cannot say that we have dealt with all of those. However, this is what we got after grinding for almost a year.

Contributed equally with the first author. In this work we attempt to come up with generalisation of dynamic cells in video segmentation, and instead of manually designing contextual blocks that connect per-frame outputs, we propose a neural architecture search solution, where the choice of operations together with their sequential arrangement are being predicted by a separate neural network.