1st Place Solution of LVIS Challenge 2020: A Good Box is not a Guarantee of a Good Mask
Not on the map. Why?
Abstract (arXiv)
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 distribution and high quality instance segmentation mask. We adopt a two-stage training pipeline. In the first stage, we incorporate EQL and self-training to learn generalized representation. In the second stage, we utilize Balanced GroupSoftmax to promote the classifier, and propose a novel proposal assignment strategy and a new balanced mask loss for mask head to get more precise mask predictions. Finally, we achieve 41.5 and 41.2 AP on LVIS v1.0 val and test-dev splits respectively, outperforming the baseline based on X101-FPN-MaskRCNN by a large margin.
Key facts. Cited by 15 papers in the map (data through 7 Oct 2026). First posted on arXiv 3 Sep 2020.
Builds on
No reference list yet.
Built on by
Citing papers per year: 2021: 6; 2022: 6; 2023: 1; 2024: 1; 2025: 1.
15 papers.
Sort:
By area:
- 2022EVA: Exploring the Limits of Masked Visual Representation Learning at Scalecited by 685
- 2022Detecting Twenty-thousand Classes using Image-level Supervisioncited by 613
- 2022MeMViT: Memory-Augmented Multiscale Vision Transformer for Efficient Long-Term Video Recognitioncited by 189
- 2021Learning Open-World Object Proposals without Learning to Classifycited by 109
- 2022X-Paste: Revisiting Scalable Copy-Paste for Instance Segmentation using CLIP and StableDiffusioncited by 49
- 2021RefineMask: Towards High-Quality Instance Segmentation with Fine-Grained Featurescited by 47
- 2021On Model Calibration for Long-Tailed Object Detection and Instance Segmentationcited by 39
- 2023MosaicFusion: Diffusion Models as Data Augmenters for Large Vocabulary Instance Segmentationcited by 35
- 2021MosaicOS: A Simple and Effective Use of Object-Centric Images for Long-Tailed Object Detectioncited by 33
- 2022Proper Reuse of Image Classification Features Improves Object Detectioncited by 20
- 2024DiverGen: Improving Instance Segmentation by Learning Wider Data Distribution with More Diverse Generative Datacited by 15
- 2021Zero-shot Object Detection Through Vision-Language Embedding Alignmentcited by 8
3 more
- 2025InstaDA: Augmenting Instance Segmentation Data with Dual-Agent System3 Sep 2025
- 2024DiverGen: Improving Instance Segmentation by Learning Wider Data Distribution with More Diverse Generative Data16 May 2024
- 2023MosaicFusion: Diffusion Models as Data Augmenters for Large Vocabulary Instance Segmentation22 Sep 2023
- 2022X-Paste: Revisiting Scalable Copy-Paste for Instance Segmentation using CLIP and StableDiffusion7 Dec 2022
- 2022EVA: Exploring the Limits of Masked Visual Representation Learning at Scale14 Nov 2022
- 2022The Equalization Losses: Gradient-Driven Training for Long-tailed Object Recognition11 Oct 2022
- 2022Proper Reuse of Image Classification Features Improves Object Detection1 Apr 2022
- 2022MeMViT: Memory-Augmented Multiscale Vision Transformer for Efficient Long-Term Video Recognition20 Jan 2022
- 2022Detecting Twenty-thousand Classes using Image-level Supervision7 Jan 2022
- 2021Zero-shot Object Detection Through Vision-Language Embedding Alignment24 Sep 2021
- 2021Learning Open-World Object Proposals without Learning to Classify15 Aug 2021
- 2021On Model Calibration for Long-Tailed Object Detection and Instance Segmentation5 Jul 2021
3 more
- 2021Content-Augmented Feature Pyramid Network with Light Linear Spatial Transformers for Object Detection20 May 2021
- 2021RefineMask: Towards High-Quality Instance Segmentation with Fine-Grained Features17 Apr 2021
- 2021MosaicOS: A Simple and Effective Use of Object-Centric Images for Long-Tailed Object Detection17 Feb 2021
- 2022EVA: Exploring the Limits of Masked Visual Representation Learning at Scalecited by 685
- 2022Detecting Twenty-thousand Classes using Image-level Supervisioncited by 613
- 2022MeMViT: Memory-Augmented Multiscale Vision Transformer for Efficient Long-Term Video Recognitioncited by 189
- 2021Learning Open-World Object Proposals without Learning to Classifycited by 109
- 2022X-Paste: Revisiting Scalable Copy-Paste for Instance Segmentation using CLIP and StableDiffusioncited by 49
- 2021RefineMask: Towards High-Quality Instance Segmentation with Fine-Grained Featurescited by 47
- 2021On Model Calibration for Long-Tailed Object Detection and Instance Segmentationcited by 39
- 2021MosaicOS: A Simple and Effective Use of Object-Centric Images for Long-Tailed Object Detectioncited by 33
- 2022Proper Reuse of Image Classification Features Improves Object Detectioncited by 20
- 2024DiverGen: Improving Instance Segmentation by Learning Wider Data Distribution with More Diverse Generative Datacited by 15
- 2021Zero-shot Object Detection Through Vision-Language Embedding Alignmentcited by 8
- 2022The Equalization Losses: Gradient-Driven Training for Long-tailed Object Recognitioncited by 8
12 most cited of 13 All 13 (JSON)
- 2024DiverGen: Improving Instance Segmentation by Learning Wider Data Distribution with More Diverse Generative Data16 May 2024
- 2022X-Paste: Revisiting Scalable Copy-Paste for Instance Segmentation using CLIP and StableDiffusion7 Dec 2022
- 2022EVA: Exploring the Limits of Masked Visual Representation Learning at Scale14 Nov 2022
- 2022The Equalization Losses: Gradient-Driven Training for Long-tailed Object Recognition11 Oct 2022
- 2022Proper Reuse of Image Classification Features Improves Object Detection1 Apr 2022
- 2022MeMViT: Memory-Augmented Multiscale Vision Transformer for Efficient Long-Term Video Recognition20 Jan 2022
- 2022Detecting Twenty-thousand Classes using Image-level Supervision7 Jan 2022
- 2021Zero-shot Object Detection Through Vision-Language Embedding Alignment24 Sep 2021
- 2021Learning Open-World Object Proposals without Learning to Classify15 Aug 2021
- 2021On Model Calibration for Long-Tailed Object Detection and Instance Segmentation5 Jul 2021
- 2021Content-Augmented Feature Pyramid Network with Light Linear Spatial Transformers for Object Detection20 May 2021
- 2021RefineMask: Towards High-Quality Instance Segmentation with Fine-Grained Features17 Apr 2021
12 newest of 13 All 13 (JSON)