DL3DV-10K

A large-scale real-world dataset for deep learning 3D vision research

CommonProductOthersDataset3D Vision

DL3DV-10K is a large-scale real-world dataset containing over 10,000 high-quality videos. Each video is manually annotated with key scene points and complexity, and also provides camera pose, NeRF depth estimation, point clouds, and 3D meshes. The dataset can be used for general NeRF research, scene consistency tracking, visual language models, and other computer vision studies.

Best AI Websites & Tools

DL3DV-10K

DL3DV-10K Alternatives

ZeroNVS — 360-Degree Full Scene Generation

Instancy — Integrated Learning Management System

In3D — Quickly generate realistic full-body 3D avatars using your phone camera.

Scale — Accelerate AI application development

GVHMR — Human motion recovery based on gravity perspective coordinates to restore world positioning.

Spectacles — AR glasses developed by Snap Inc., offering enhanced reality experiences.

World Labs — Build large world models, perceive, generate, and interact with the 3D world.

GameGen-O — A diffusion transformation model designed for generating open-world video games.

OmniRe — High-fidelity dynamic urban scene reconstruction technology

CSGO — Application of content style synthesis in text-to-image generation.

Shusheng: Skyline LandMark — A city-scale NeRF-based immersive 3D model offering an engaging experience.

Kolors Virtual Try-On — AI-powered virtual try-on experience

MedTrinity-25M — A large-scale multimodal medical dataset

Data-Juicer — A one-stop data processing system that provides high-quality data for large language models.

avp_teleoperate — Remote control of the humanoid robot Unitree H1_2 using Apple Vision Pro.

ControlMM — Full-body motion generation framework supporting multimodal control

MINT-1T — A multimodal dataset comprising one trillion tokens and 3.4 billion images.

SA-V Dataset — Video dataset for training general object segmentation models.

Segment Anything Model 2 — A foundational model for visual segmentation of images and videos.

HoloDreamer — Generate holographic 3D panoramic worlds from text descriptions.

Aiuni — Explore the 3D virtual world and experience the Dreamer's Universe.

DCLM-baseline — High-performance language model benchmark dataset

UltraEdit — Large-Scale Image Editing Dataset

EgoGaussian — 3D Scene Reconstruction and Dynamic Object Tracking Technology

Ario — Leveraging augmented reality technology, Ario provides an easy-to-use knowledge-sharing tool for frontline workers, delivering information at the right time, place, and context.

PixelProse — A large-scale image captioning dataset providing over 16M synthetic image descriptions.

WonderWorld — Generates interactive 3D scenes from a single image

emo-visual-data — Emoji Visual Annotation Dataset

IllumiNeRF — 3D relighting technology, without the need for inverse rendering