CompositionEvidence: Controlling Geometric Evidence for Cinematic Composition Reasoning with Frozen Vision-Language Models
cs.CV
Submitted: 2026-11-07
Updated: 2026-11-07
Code: https://github.com/Cyber-Yichen/composition-evidence-artifact
Terminology
Sources
- Qwen2.5-VL Technical Report
- Geometry-Guided Camera Motion Understanding in VideoLLMs
- MovieNet: A Holistic Dataset for Movie Understanding
- LLaVA-OneVision: Easy Visual Task Transfer
- ShotBench: Expert-Level Cinematic Understanding in Vision-Language Models
- Grounding DINO: Marrying DINO with Grounded Pre-Training for Open-Set Object Detection
- A Unified Framework for Shot Type Classification Based on Subject Centric Lens
- CineTechBench: A Benchmark for Cinematographic Technique Understanding and Generation
- RefineShot: Rethinking Cinematography Understanding with Foundational Skill Evaluation
- CamReasoner: Reinforcing Camera Movement Understanding via Structured Spatial Reasoning
- Depth Anything V2
- PhotoFramer: Multi-modal Image Composition Instruction
- Predicting Camera Pose from Perspective Descriptions for Spatial Reasoning
- InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Models
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