28 citations · 156 across the 18 of their papers we have counts for
8 papers · 1 filter
Equalization Loss v2: A New Gradient Balance Approach for Long-tailed Object Detection
Jingru Tan, Xin Lu, Gang Zhang +2
Recently proposed decoupled training methods emerge as a dominant paradigm for long-tailed object detection. But they require an extra fine-tuning stage, and the disjointed optimiz…
Dynamic Graph: Learning Instance-aware Connectivity for Neural Networks
Kun Yuan, Quanquan Li, Dapeng Chen +2
One practice of employing deep neural networks is to apply the same architecture to all the input instances. However, a fixed architecture may not be representative enough for data…
MimicDet: Bridging the Gap Between One-Stage and Two-Stage Object Detection
Xin Lu, Quanquan Li, Buyu Li +1
Modern object detection methods can be divided into one-stage approaches and two-stage ones. One-stage detectors are more efficient owing to straightforward architectures, but the…
1st Place Solution of LVIS Challenge 2020: A Good Box is not a Guarantee of a Good Mask
Jingru Tan, Gang Zhang, Hanming Deng +4
This article introduces the solutions of the team lvisTraveler for LVIS Challenge 2020. In this work, two characteristics of LVIS dataset are mainly considered: the long-tailed dis…
Learning Connectivity of Neural Networks from a Topological Perspective
Kun Yuan, Quanquan Li, Jing Shao +1
Seeking effective neural networks is a critical and practical field in deep learning. Besides designing the depth, type of convolution, normalization, and nonlinearities, the topol…
DMCP: Differentiable Markov Channel Pruning for Neural Networks
Shaopeng Guo, Yujie Wang, Quanquan Li +1
Recent works imply that the channel pruning can be regarded as searching optimal sub-structure from unpruned networks. However, existing works based on this observation require tra…
Equalization Loss for Long-Tailed Object Recognition
Jingru Tan, Changbao Wang, Buyu Li +4
Object recognition techniques using convolutional neural networks (CNN) have achieved great success. However, state-of-the-art object detection methods still perform poorly on larg…
Residual Knowledge Distillation
Mengya Gao, Yujun Shen, Quanquan Li +1
Knowledge distillation (KD) is one of the most potent ways for model compression. The key idea is to transfer the knowledge from a deep teacher model (T) to a shallower student (S)…