Segment-Any-Anomaly
Image anomaly detector
An implementation of a method to identify anomalies in images without needing training data
Official implementation of "Segment Any Anomaly without Training via Hybrid Prompt Regularization (SAA+)".
753 stars
10 watching
75 forks
Language: Jupyter Notebook
last commit: almost 3 years agoanomaly-detectionfoundationzero-shot
Related projects:
| Repository | Description | Stars |
|---|---|---|
| Detects anomalies in hyperspectral images using an autoencoder-based approach | 45 | |
| A package providing functions to decompose and detect anomalies in time series data | 339 | |
| An investigation into the relationship between misleading images and hallucinations in large language models | 8 | |
| Develops a system to detect, segment, and rank camouflaged objects in images. | 74 | |
| A toolkit for rule-based and unsupervised anomaly detection in time series data | 1,108 | |
| A MATLAB implementation of Support Vector Data Description (SVDD) for anomaly detection and classification. | 78 | |
| Develops a neural network architecture for object detection and instance segmentation without labeled data | 52 | |
| An anomaly detection framework for flexible and efficient metric analysis | 280 | |
| Provides Suricata IDS alert rules for detecting network anomalies | 154 | |
| An object detection framework that uses extreme and center points to detect objects in images. | 1,032 | |
| Develops an object segmentation algorithm to detect camouflaged objects in images with varying backgrounds and contexts. | 20 | |
| Develops an algorithm to segment camouflaged objects in images using a neural network architecture called Anabranch Network. | 4 | |
| This repository provides code for training a model to detect anomalies in graph data using pattern mining and feature learning. | 40 | |
| This project provides code and pre-trained models for object segmentation in videos using a spatial propagation network. | 144 | |
| Automated image segmentation of road and vehicle features from video data using the U-Net algorithm | 15 |