Research map / Detection and segmentation / CNN features
SipMask: Spatial Information Preservation for Fast Image and Video Instance Segmentation
Jiale Cao, Rao Muhammad Anwer, Hisham Cholakkal, Fahad Shahbaz Khan, Yanwei Pang, Ling Shao
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add spatial preservation (SP) module (ablation) | YOLACT-like baseline · COCO minival (ablation) | mask AP: 31.2 -> 33.4 | Improved |
| add contextual basis masks (CBM) on top of SP (ablation) | SipMask · COCO minival (ablation) | mask AP: 33.4 -> 33.8 | Improved |
| add mask alignment weighting loss on top of SP and CBM (ablation) | SipMask · COCO minival (ablation) | mask AP: 33.8 -> 34.3 | Improved |
| add spatial coefficients alone (ablation) | baseline · COCO minival (ablation) | mask AP: 31.2 -> 32.9 | Improved |
| add feature alignment alone (ablation) | baseline · COCO minival (ablation) | mask AP: 31.2 -> 31.7 | Improved |
| add contextual basis masks alone (ablation) | baseline · COCO minival (ablation) | mask AP: 31.2 -> 31.9 | Improved |
| add weighting loss alone (ablation) | baseline · COCO minival (ablation) | mask AP: 31.2 -> 32.0 | Improved |
| split box into 2x2 sub-regions instead of 1x1 (ablation) | SipMask SP module · COCO minival (ablation) | mask AP: 31.2 -> 32.9 | Improved |
| increase sub-regions from 2x2 to 3x3 (ablation) | SipMask SP module · COCO minival (ablation) | mask AP: 32.9 -> 33.1 | No clear effect |
| increase sub-regions from 3x3 to 4x4 (ablation) | SipMask SP module · COCO minival (ablation) | mask AP: 33.1 -> 33.1 | No clear effect |
| weight mask loss by both class score and overlap instead of class score only (ablation) | mask alignment weighting · COCO minival (ablation) | mask AP: 31.8 -> 32.0 | Improved |
| weight mask loss by both class score and overlap instead of overlap only (ablation) | mask alignment weighting · COCO minival (ablation) | mask AP: 31.7 -> 32.0 | Improved |
Later papers that compared with this method
2 later accepted papers report SipMask as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- SG-Net: Spatial Granularity Network for One-Stage Video Instance Segmentation (CVPR 2021 · better on 2 of 4 shared results)
- Track To Detect and Segment: An Online Multi-Object Tracker (CVPR 2021 · better on 1 of 1 shared result)
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 4 on most shared results.
- TensorMask — TensorMask: A Foundation for Dense Object Segmentation (ICCV 2019) · better on 2 of 2 shared results
- D2Det — D2Det: Towards High Quality Object Detection and Instance Segmentation (CVPR 2020) · better on 0 of 1 shared result
- PolarMask — PolarMask: Single Shot Instance Segmentation With Polar Representation (CVPR 2020) · better on 1 of 1 shared result
- PANet — PANet: Few-Shot Image Semantic Segmentation With Prototype Alignment (ICCV 2019) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- SipMaskv2: Enhanced Fast Image and Video Instance Segmentation (TPAMI 2022 · similarity 0.98)
- Mask-Free Video Instance Segmentation (CVPR 2023 · similarity 0.90)
- SOLOv2: Dynamic and Fast Instance Segmentation (NeurIPS 2020 · similarity 0.90)
- PolarMask: Single Shot Instance Segmentation With Polar Representation (CVPR 2020 · similarity 0.90)
- Real-Time Instance Segmentation With Discriminative Orientation Maps (ICCV 2021 · similarity 0.90)
- BlendMask: Top-Down Meets Bottom-Up for Instance Segmentation (CVPR 2020 · similarity 0.90)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
