OBA-Research/VAAS
VAAS is an inference-first, research-driven library for image integrity analysis. It integrates Vision Transformer Attention Mechanisms with patch-level self-consistency analysis to enable fine-grained localization and detection of visual inconsistencies across diverse image analysis tasks.
This tool helps researchers, digital forensics experts, and content evaluators analyze images to detect subtle visual inconsistencies that might indicate manipulation or anomalies. You input an image, and it outputs a set of numerical scores indicating global and local anomaly intensity, along with a visual heatmap highlighting suspicious areas. This helps you identify images that might have been altered or generated.
Available on PyPI.
Use this if you need to determine the integrity of an image and pinpoint specific regions that appear anomalous or potentially manipulated.
Not ideal if you need a simple 'yes/no' binary decision on image authenticity without detailed anomaly scoring or visual explanations.
Stars
25
Forks
3
Language
Python
License
MIT
Category
Last pushed
Feb 26, 2026
Commits (30d)
0
Dependencies
7
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/computer-vision/OBA-Research/VAAS"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Related tools
deepmancer/clip-object-detection
Zero-shot object detection with CLIP, utilizing Faster R-CNN for region proposals.
ABaldrati/CLIP4Cir
[ACM TOMM 2023] - Composed Image Retrieval using Contrastive Learning and Task-oriented...
kyegomez/CLIPQ
A simple implementation of a CLIP that splits up an image into quandrants and then gets the...
IvanAer/G-Universal-CLIP
4th place solution for the Google Universal Image Embedding Kaggle Challenge. Instance-Level...
joanrod/ocr-vqgan
OCR-VQGAN, a discrete image encoder (tokenizer and detokenizer) for figure images in...