LVIS Challenge Track Technical Report 1st Place Solution: Distribution Balanced and Boundary Refinement for Large Vocabulary Instance Segmentation
By Weifu Fu, Congchong Nie, Ting Sun, J. Liu, Tianliang Zhang, Yong Liu. 2021
6 citations recorded by the source corpus.
This report introduces the technical details of the team FuXi-Fresher for LVIS Challenge 2021. Our method focuses on the problem in following two aspects: the long-tail distribution and the segmentation quality of mask and boundary. Based on the advanced HTC instance segmentation algorithm, we connect transformer backbone(Swin-L) through composite connections inspired by CBNetv2 to enhance the baseline results. To alleviate the problem of long-tail distribution, we design a Distribution Balanced method which includes dataset balanced and loss function balaced modules. Further, we use a Mask and Boundary Refinement method composed with mask scoring and refine-mask algorithms to improve the segmentation quality. In addition, we are pleasantly surprised to find that early stopping combined with EMA method can achieve a great improvement. Finally, by using multi-scale testing and increasing the upper limit of the number of objects detected per image, we achieved more than 45.4% boundary AP on the val set of LVIS Challenge 2021. On the test data of LVIS Challenge 2021, we rank 1st and achieve 48.1% AP. Notably, our APr 47.5% is very closed to the APf 48.0%.
Subfields: Computer Vision and Pattern Recognition
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Computer Vision and Pattern Recognition
- Xinggang Wang
- Luc Van Gool
- Dacheng Tao
- Chunhua Shen
- Ming-Hsuan Yang
- Alan L. Yuille
- Trevor Darrell
- Jitendra Malik
Citations in the corpus
- Exploring Plain Vision Transformer Backbones for Object Detection (2022, methodology): “7 competition winner 2021 [17]†, baseline 21K, sup 43.”
- EVA: Exploring the Limits of Masked Visual Representation Learning at Scale (2022, background): “best” refers to the best individual model / result in each benchmark (a: DINO [121], b: ViTDet-H [60], c: Mask DINO [57], d: 2021 competition 1 [35]) “∆↓”: the performance gap between LVIS and COCO (the lower the better).”