InstAP: Instance-Aware Vision-Language Pre-Train for Spatial-Temporal Understanding
Ashutosh Kumar, Rajat Saini, Jingjing Pan, Mustafa Erdogan, Mingfang Zhang +3 more
Abstract
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...
Results & Benchmarks
| Task | Dataset | Metric | Value |
|---|---|---|---|
| Retrieval | Baseline | InstVL-1K | 59.10 |
| Retrieval | + Instance temperature | InstVL-1K | 67.19 |
Hardware Requirements
- Expect multi-day setup/compute for meaningful reproduction based on current guidance.
Best Implementation
Maintained implementation evidence is not confirmed for this paper yet.
Use the Implementation Status and Reproduction Path sections below for the current action plan.
Reproduction Path
Follow this baseline workflow to decide if this paper is worth immediate prototyping.
- 1
Use the paper and benchmark evidence to scope a baseline reproduction plan.
- 2
Track assumptions and missing details in an experiment log before coding.
Framework baselines
- Hugging Face Transformers training guide
Modern transformer training baseline.
- PyTorch nn.Transformer docs
Reference transformer building block implementation.
Additional Implementations
No additional verified repositories beyond the primary recommendation.
Hugging Face Artifacts
No trustworthy direct or curated related Hugging Face artifacts were found yet.
Continue with targeted Hugging Face searches: