InstAP: Instance-Aware Vision-Language Pre-Train for Spatial-Temporal Understanding
Current vision-language pre-training (VLP) paradigms excel at global scene understanding but struggle with instance-level reasoning due to global-only supervision. We introduce InstAP, an Instance-Aware Pre-training framework that jointly optimizes global vision-text alignment and fine-grained, instance-level contrastive alignment by grounding textual mentions to specific spatial-temporal regions. To support this, we present InstVL, a large-scale dataset (2 million images, 50,000 videos) with dual-granularity annotations: holistic scene captions and dense, grounded instance descriptions. On the InstVL benchmark, InstAP substantially outperforms existing VLP models on instance-level retrie...
| Authors | Ashutosh Kumar, Rajat Saini, Jingjing Pan et al. |
|---|---|
| Primary category | cs.CV |
| Published | 9 April 2026 |
| Last updated at source | 9 April 2026 |
Popularity, last 3 days
0 now. Counts as reported by the source, snapshotted daily by Silytics.
Connected in the knowledge graph
- wovenbytoyota-vai/InstVLDataset · cited by
Models and datasets referencing this paper
- wovenbytoyota-vai/InstVLDataset
Silytics shows public metadata from arXiv and links to the original. Information may be incomplete or out of date. Last refreshed 19 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.