Research map / Detection and segmentation / Open-set and few-shot detection
T-Rex2++: Toward Generic Object Perception via Text-Visual Prompt Synergy
Qing Jiang, Feng Li, Zhaoyang Zeng, Tianhe Ren, Shilong Liu, Lei Zhang
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 contrastive alignment between text and visual prompts (ablation) | T-Rex2 text prompt · LVIS-Val zero-shot | AP: 32.2 -> 34.8 (+2.6) | Improved |
| add contrastive alignment between text and visual prompts (ablation) | T-Rex2 visual-G prompt · LVIS-Val zero-shot | AP: 30.2 -> 34.9 (+4.7) | Improved |
| add contrastive alignment between text and visual prompts (ablation) | T-Rex2 text prompt · COCO-Val zero-shot | AP: 44.4 -> 45.8 (+1.4) | Improved |
| add contrastive alignment between text and visual prompts (ablation) | T-Rex2 visual-G prompt · COCO-Val zero-shot | AP: 38.7 -> 38.8 (+0.1) | No clear effect |
| train text and visual prompts jointly without alignment instead of visual only (ablation) | T-Rex2 visual-G prompt · COCO-Val zero-shot | AP: 14.0 -> 38.7 | Improved |
| naive joint training of both prompts instead of text only (ablation) | T-Rex2 text prompt · COCO-Val zero-shot | AP: 46.4 -> 44.4 | Hurt |
| increase number of visual prompts per category from 1 to 16 (ablation) | T-Rex2 visual-G prompt · COCO-Val zero-shot | AP: 29.2 -> 38.8 | Improved |
| increase number of visual prompts per category from 16 to 64 (ablation) | T-Rex2 visual-G prompt · COCO-Val zero-shot | AP: 38.8 -> 41.4 | Improved |
| combine text and visual-G prompts at inference (ablation) | T-Rex2 · LVIS-Val zero-shot | AP: 34.8 -> 37.0 | Improved |
| combine text and visual-G prompts at inference (ablation) | T-Rex2 · COCO-Val zero-shot | AP: 45.8 -> 42.5 | Hurt |
| add Bamboo dataset to text prompt training (ablation) | T-Rex2-T text prompt · LVIS-Minival zero-shot | AP: 34.9 -> 38.7 (+3.8AP) | Improved |
| add Bamboo dataset to text prompt training (ablation) | T-Rex2-T text prompt · COCO-Val zero-shot | AP: 46.1 -> 45.7 (-0.4AP) | Hurt |
Closest papers by meaning
- Just a Few Glances: Open-Set Visual Perception with Image Prompt Paradigm (AAAI 2025 · similarity 0.90)
- DoGA: Enhancing Grounded Object Detection via Grouped Pre-Training with Attributes (AAAI 2025 · similarity 0.90)
- Revisiting Few-Shot Object Detection with Vision-Language Models (NeurIPS 2024 · similarity 0.89)
- Towards Single-Source Domain Generalized Object Detection via Causal Visual Prompts (NeurIPS 2025 · similarity 0.88)
- CP-DETR: Concept Prompt Guide DETR Toward Stronger Universal Object Detection (AAAI 2025 · similarity 0.88)
- Roboflow100-VL: A Multi-Domain Object Detection Benchmark for Vision-Language Models (NeurIPS 2025 · similarity 0.88)
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.
