Video-Depth-Estimation-Rankings-and-Stereo-Video-Conversion-Rankings
We look forward to models based on MiniMax H3. ChronoDepth Depth Any Video Depth Anything Depth Pro DepthCrafter DreamStereo DVD Elastic3D Eye2Eye FFN GemDepth
What is Video-Depth-Estimation-Rankings-and-Stereo-Video-Conversion-Rankings?
We look forward to models based on MiniMax H3. ChronoDepth Depth Any Video Depth Anything Depth Pro DepthCrafter DreamStereo DVD Elastic3D Eye2Eye FFN GemDepth GRT HairGuard InfiniDepth MoGe M2SVid NVDS Pixel-Perfect Video Depth Restereo StereoCrafter StereoEngine StereoPilot StereoWorld SVG UniDepth UniK3D Video Depth Anything ViGeo αDepth
Overview of Video-Depth-Estimation-Rankings-and-Stereo-Video-Conversion-Rankings
What it is
This repository provides rankings and curated lists for video depth estimation and stereo video conversion. It includes tables of recent stereo video conversion methods (from April 2025 to October 2026) with their backbones and submission dates, as well as synthetic RGB-D image and video datasets for training high-definition video depth estimation models. The repository also presents benchmark results for video depth estimation (e.g., 170-frame ScanNet: TAE, 500-frame Bonn RGB-D Dynamic: δ1) and single image depth estimation (Synth4K: δ1).
Key features
- Lists stereo video conversion methods from the last 19 months (1 April 2025 to 12 October 2026) with backbone and arXiv submission date.
- Curates synthetic RGB-D image datasets (SynthHuman, MegaSynth) and video datasets (e.g., SynthVerse, PhysInOne, SDG-SynHuman) for training HD video depth estimation models.
- Provides rankings for video depth estimation on ScanNet and Bonn RGB-D Dynamic, and for single image depth estimation on Synth4K.
- Encourages researchers to develop new stereo video conversion models based on MiniMax H3.
- Includes appendices with notes on dataset selection, metrics, and lists of research papers.
Use cases
Researchers can use this repository to compare the state of the art in stereo video conversion and video depth estimation, select appropriate synthetic datasets for training or testing, and identify benchmark results for their own models.
Getting started
See the repository README for details.
License and pricing
The license is not stated. The repository is listed as open source.
Based on the project's public README and repository metadata.