Research map / Detection and segmentation / Open-set and few-shot detection
X-SAM: From Segment Anything to Any Segmentation
Hao Wang, Limeng Qiao, Zequn Jie, Zhijian Huang, Chengjian Feng, Qingfang Zheng, Lin Ma, Xiangyuan Lan, Xiaodan Liang
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 mixed co-training across tasks (ablation) | X-SAM · A150-OV | PQ: 16.4 -> 22.4 | Improved |
| add mixed co-training across tasks (ablation) | X-SAM · Reason-Val | gIoU: 48.2 -> 57.1 | Improved |
| add mixed co-training across tasks (ablation) | X-SAM · COCO-Pan | PQ: 55.3 -> 54.5 | Hurt |
| add SAM segmentation encoder next to ViT (ablation) | X-SAM · GCG-Val | mIoU: 64.8 -> 69.4 | Improved |
| add Swin (M2F-init) segmentation encoder next to ViT (ablation) | X-SAM · GCG-Val | mIoU: 64.8 -> 62.5 | Hurt |
| add Swin (M2F-init) segmentation encoder next to ViT (ablation) | X-SAM · COCO-VGD | AP: 40.7 -> 48.6 | Improved |
| add stage 1 segmentor fine-tuning (ablation) | X-SAM · COCO-Pan | PQ: 45.2 -> 54.5 | Improved |
| add stage 2 alignment pre-training (ablation) | X-SAM · GCG-Val | mIoU: 65.4 -> 69.4 | Improved |
| add stage 2 alignment pre-training (ablation) | X-SAM · COCO-Pan | PQ: 54.5 -> 54.7 | No clear effect |
| replace SAM decoder with Mask2Former decoder (ablation) | X-SAM · segmentor fine-tuning (12 epochs) | PQ: 40.9 -> 50.1 | Improved |
| use convolution connector instead of MLP connector (ablation) | X-SAM · segmentor fine-tuning (12 epochs) | PQ: 50.1 -> 50.3 | Improved |
| add multi-scale features in connector (ablation) | X-SAM · segmentor fine-tuning (12 epochs) | PQ: 50.3 -> 51.6 | Improved |
Later papers that compared with this method
2 later accepted papers report X-SAM as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- SetCon: Towards Open-Ended Referring Segmentation via Set-Level Concept Prediction (NeurIPS 2026 · better on 1 of 1 shared result)
- Conversational Image Segmentation: Grounding Abstract Concepts with Scalable Supervision (CVPR 2026 · better on 0 of 1 shared result)
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- PSALM — PSALM: Pixelwise Segmentation with Large Multi-modal Model (ECCV 2024) · better on 5 of 5 shared results
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
- ARGenSeg: Image Segmentation with Autoregressive Image Generation Model (NeurIPS 2025 · similarity 0.90)
- Rethinking Interactive Image Segmentation with Low Latency High Quality and Diverse Prompts (CVPR 2024 · similarity 0.89)
- SAM-CP: Marrying SAM with Composable Prompts for Versatile Segmentation (ICLR 2025 · similarity 0.89)
- OpenWorldSAM: Extending SAM2 for Universal Image Segmentation with Language Prompts (NeurIPS 2025 · similarity 0.89)
- Segment Anything without Supervision (NeurIPS 2024 · similarity 0.89)
- SAM 2: Segment Anything in Images and Videos (ICLR 2025 · similarity 0.89)
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.
