Virtual Task Prompting for Multi-Task Scene Understanding
cs.CV
Submitted: 2026-11-08
Updated: 2026-11-08
Project page: https://jacky1128.github.io/VTP
Terminology
Sources
- Qwen3-VL Technical Report
- Embodied Multimodal Multitask Learning
- Virtual Width Networks
- UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction
- DINOv3
- ViT-5: Vision Transformers for The Mid-2020s
- mHC: Manifold-Constrained Hyper-Connections
Related papers
- TEA: Text Encoder Alignment for Robust Concept Erasure in Text-to-Image Models
- Do Egocentric Video-Language Models Capture Both Hand- and Object-Centric Cues?
- Aura: Consistent Multi-Subject Video Generation via VLM-Grounded Semantic Alignment
- PVRF: All-in-one Adverse Weather Removal via Prior-modulated and Velocity-constrained Rectified Flow
- DLWM: Diverse Latent World Models for Efficient Multimodal Reasoning
- Out of Sight, Still in Mind: Token Compression for Omni-LLMs