UMI-VQA-8M: 8 Million Visual Q&A Samples for Wrist-Mounted Fisheye Observations

UMI-VQA-8M is a large-scale visual question answering dataset built for UMI-style wrist-mounted fisheye observations. It contains 8 million visual question-answering samples and provides visual-language supervision for UMI observation scenarios. The dataset was created by TeleEmbodied and was last updated on the Hugging Face platform in June 2026.

Use Cases

• Training visual question answering models based on fisheye observations mentioned in the description.
• Developing robotics perception systems based on the UMI-style wrist-mounted observation scenario.
• Benchmarking multimodal AI models on tasks grounded in physics-validated visual data.

Strengths

• Contains 8 million visual question-answering samples, indicating a large scale.
• Specifically designed for UMI-style wrist-mounted fisheye observations, providing domain relevance.
• Consists of two complementary parts, including a real-world subset, as mentioned in the description.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

TeleEmbodied on Hugging Face.

Collection Method

Likely contains real-world and simulated visual question-answering data for UMI observation scenarios.

Freshness

Last updated 2026-06-04 08:25:30; freshness should be verified.