pytorch-grad-cam and Grad-CAM

The pytorch-grad-cam library provides the core implementation that powers the Grad-CAM web demo, making them ecosystem siblings where one is the underlying technical framework and the other is a user-friendly interface for the same visualization technique.

pytorch-grad-cam
60
Established
Grad-CAM
37
Emerging
Maintenance 0/25
Adoption 13/25
Maturity 25/25
Community 22/25
Maintenance 0/25
Adoption 9/25
Maturity 8/25
Community 20/25
Stars: 12,682
Forks: 1,694
Downloads:
Commits (30d): 0
Language: Python
License: MIT
Stars: 110
Forks: 25
Downloads:
Commits (30d): 0
Language: HTML
License:
Stale 6m
No License Stale 6m No Package No Dependents

About pytorch-grad-cam

jacobgil/pytorch-grad-cam

Advanced AI Explainability for computer vision. Support for CNNs, Vision Transformers, Classification, Object detection, Segmentation, Image similarity and more.

This helps data scientists, machine learning engineers, and researchers understand why their computer vision AI models make specific decisions. You input a trained image classification, object detection, or segmentation model, and it outputs visual heatmaps showing the exact regions of an image that influenced the model's prediction. This allows users to diagnose model errors, build trust in AI systems, and improve model performance.

AI-explainability computer-vision model-debugging machine-learning-operations deep-learning-research

About Grad-CAM

Cloud-CV/Grad-CAM

:rainbow: :camera: Gradient-weighted Class Activation Mapping (Grad-CAM) Demo

This tool helps researchers and analysts understand why an AI model made a specific prediction when analyzing images. It takes an image that has been processed by a Convolutional Neural Network (CNN) and outputs a 'heat map' overlaying the original image, highlighting the regions that were most important for the model's decision. This is ideal for anyone who needs to interpret or explain the reasoning behind an AI's image-based predictions.

AI-explainability computer-vision image-analysis model-interpretation visual-question-answering

Scores updated daily from GitHub, PyPI, and npm data. How scores work