Open Vision Banana: A Fully Open Recipe for Generative Visual Perception
cs.CV
Submitted: 2026-11-11
Updated: 2026-11-11
Terminology
Sources
- Qwen3-VL Technical Report
- MiDaS v3.1 -- A Model Zoo for Robust Monocular Relative Depth Estimation
- Virtual KITTI 2
- Image Generators are Generalist Vision Learners
- Vision as Unified Multimodal Generation
- GenAI-Bench: Evaluating and Improving Compositional Text-to-Visual Generation
- Seg-Zero: Reasoning-Chain Guided Segmentation via Cognitive Reinforcement
- Public Domain 12M: A Highly Aesthetic Image-Text Dataset with Novel Governance Mechanisms
- SPair-71k: A Large-scale Benchmark for Semantic Correspondence
- Pico-Banana-400K: A Large-Scale Dataset for Text-Guided Image Editing
- DINO-X: A Unified Vision Model for Open-World Object Detection and Understanding
- Steerable Visual Representations
- The Replica Dataset: A Digital Replica of Indoor Spaces
- DIODE: A Dense Indoor and Outdoor DEpth Dataset
- HyperSeg: Towards Universal Visual Segmentation with Large Language Model
- Synscapes: A Photorealistic Synthetic Dataset for Street Scene Parsing
- Depth Anything V2
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