Research map / Video / Video segmentation and tracking
Temporal-Emerged Prompting for Segment Anything in Multiframe Infrared Small Target Detection
Yinghui Xing, Donghao Chu, Shizhou Zhang, Di Xu
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 temporal-emerged prompting and feature injection (TEP-SAM) over best prior multiframe method | TEP-SAM (SAM-B) · TSIRMT (All) | IoU: 65.89 -> 74.34 | Improved |
| add DETE temporal encoder (vs per-frame conv + average pooling) (ablation) | TEP-SAM · NUDT-MIRSDT | IoU: 64.84 -> 86.15 | Improved |
| add temporal feature injection (TFI) (ablation) | TEP-SAM · NUDT-MIRSDT | IoU: 50.30 -> 86.15 | Improved |
| add temporal prompt generator (TP-Gen) (ablation) | TEP-SAM · NUDT-MIRSDT | IoU: 84.86 -> 86.15 IoU; Pd 99.77 -> 99.71 | Mixed |
| add auxiliary loss on DETE (ablation) | TEP-SAM · NUDT-MIRSDT | IoU: 82.69 -> 86.15 | Improved |
| add TDM branch alone, without GLTM (ablation) | DETE · NUDT-MIRSDT | IoU: 80.30 -> 79.76 | Hurt |
| add TDM together with GLTM branches (ablation) | DETE · NUDT-MIRSDT | IoU: 85.42 -> 86.15 | Improved |
| add global temporal features alone (ablation) | DETE · NUDT-MIRSDT | IoU: 80.30 -> 85.31 | Improved |
| extend single-frame SAM-SPL with temporal pooling | SAM-SPL · NUDT-MIRSDT (All) | IoU: 47.47 -> 33.73 | Hurt |
| extend SAM-SPL with DTUM | SAM-SPL · NUDT-MIRSDT (All) | IoU: 47.47 -> 76.28 | Improved |
| use larger SAM-L instead of SAM-B encoder (ablation) | TEP-SAM · NUDT-MIRSDT (All) | IoU: 86.15 -> 86.27 | No clear effect |
Closest papers by meaning
- Motion Prior Knowledge Learning with Homogeneous Language Descriptions for Moving Infrared Small Target Detection (AAAI 2025 · similarity 0.88)
- Rethinking Infrared Small Target Detection: A Foundation Driven Efficient Paradigm (NeurIPS 2026 · similarity 0.88)
- MOCID: Motion Context and Displacement Information Learning for Moving Infrared Small Target Detection (AAAI 2025 · similarity 0.87)
- Endow SAM with Keen Eyes: Temporal-spatial Prompt Learning for Video Camouflaged Object Detection (CVPR 2024 · similarity 0.87)
- SAIST: Segment Any Infrared Small Target Model Guided by Contrastive Language-Image Pretraining (CVPR 2025 · similarity 0.85)
- Relax Image-Specific Prompt Requirement in SAM: A Single Generic Prompt for Segmenting Camouflaged Objects (AAAI 2024 · similarity 0.85)
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.
