InstVL: A Large-Scale Instance-Aware Vision-Language Dataset

InstVL is a large-scale dataset of images and videos designed for instance-aware vision-language pre-training. The dataset was created by wovenbytoyota-vai and introduced in the paper 'InstAP: Instance-Aware Vision-Language Pre-Train for Spatial-Temporal Understanding'. It was last updated on the platform in April 2026.

Use Cases

• Training instance-aware vision-language models based on the dataset's focus on fine-grained comprehension.
• Benchmarking spatial-temporal understanding capabilities in multimodal AI systems.
• Pre-training models to bridge the gap between holistic scene and instance-level understanding as described.

Strengths

• Dataset is described as 'large-scale', suggesting substantial volume.
• Focuses on the specific research area of instance-aware and spatial-temporal understanding.

Limitations

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

Provenance

Source

wovenbytoyota-vai

Collection Method

Method of data gathering is not specified in the provided description.

Time Range

Temporal coverage is not specified in the provided description.

Freshness

Last updated 2026-04-10 02:31:02; freshness should be verified.

Geography

Spatial coverage is not specified in the provided description.