Research map / Vision-language and multimodal / Robot action and navigation
Bring My Cup! Personalizing Vision-Language-Action Models with Visual Attentive Prompting
Sangoh Lee, Sangwoo Mo, Wook-Shin Han
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 Visual Attentive Prompting (mask-aligned identity tinting) | pi0 · Personalized-SIMPLER Google Robot, Track 1 | SR / CMR (%): 8.5%/10.5% -> 60.3%/89.2% | Improved |
| add VAP under variant aggregation | pi0 · Personalized-SIMPLER Google Robot, Track 2 | SR / CMR (%): SR 58.2%, CMR 87.3% | Improved |
| add VAP | pi0 · Personalized-SIMPLER WidowX Task 3 pick shaver | SR (%): 0.0 -> 71.3 | Improved |
| add VAP | pi0.5 · Personalized-VLABench Task 1 leather bag | SR (%): +36.8 points over strongest baseline | Improved |
| add VAP | pi0.5 · Real-world selection tasks | SR (%): 30.0% -> 80.0% | Improved |
| add VAP | pi0.5 · Real-world pick-and-place tasks | average SR (%): 18.8% -> 58.8% | Improved |
| add soft/hard text prompts instead of visual prompt | pi0.5 · Real-world pick-and-place tasks | average SR (%): 27.5-31.2% | No clear effect |
| use text-only hard prompts | pi0.5 · Real-world selection tasks | SR (%): 40.0-45.0% | No clear effect |
| replace mask-aligned identity tinting with points, trajectory arrows or numbered markers (ablation) | VAP · WidowX harder setting | SR (%): 83.5% -> 28-36% | Hurt |
| replace rule-based parser with GPT-4o-mini for implicit expressions (ablation) | VAP · Personalized-SIMPLER | SR (%): T1 60.3 -> 59.5; T2 75.0 -> 72.0; T3 71.3 -> 71.6; T4 92.1 -> 91.3; T5 95.0 -> 91.2; T6 75.6 -> 75.9 | No clear effect |
| add SAM2 tracking per step | VAP · runtime | time per step (s): 0.02 | No clear effect |
Closest papers by meaning
- Object-Centric Prompt-Driven Vision-Language-Action Model for Robotic Manipulation (CVPR 2025 · similarity 0.90)
- Ego: Embedding-Guided Personalization of Vision-Language Models (CVPR 2026 · similarity 0.88)
- PIVOT: Iterative Visual Prompting Elicits Actionable Knowledge for VLMs (ICML 2024 · similarity 0.88)
- Memory-Space Visual Prompting for Efficient Vision-Language Fine-Tuning (ICML 2024 · similarity 0.88)
- PERIA: Perceive, Reason, Imagine, Act via Holistic Language and Vision Planning for Manipulation (NeurIPS 2024 · similarity 0.87)
- Exploring the Limits of Vision-Language-Action Manipulation in Cross-task Generalization (NeurIPS 2025 · similarity 0.87)
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.
