Jinglianwen releases a binocular 3D first-person embodied dataset
Jinglianwen Technology released a first-person robot-training dataset with binocular RGB, time-of-flight depth and aligned action and semantic data across six real-world scenario groups.
Evidence notes
- The capture system combines 4096 by 1536 binocular RGB with an independent time-of-flight depth camera.
- Delivery includes calibration files for camera intrinsics, distortion and spatial extrinsics.
- The dataset targets visual-language-action and sim-to-real development.
Company context
AI data infrastructure company providing multimodal collection, governance, annotation, quality control and dataset operations for model developers, robotics teams and public-sector data programmes.