Flexible Feature visualization on PyTorch, for research and art 🔎 💻 🧠 🎨

Visualization tool for point cloud and feature extracted from deep learning network

An interpretability library for pytorch

This repository contains an end-to-end pipeline to train a convolutional neural network (CNN) for environmental sound classification on ESC-50. Serve the model for inference using Modal + a FastAPI endpoint. - Visualize model outputs (top predictions, input mel-spectrogram, waveform, and CNN feature maps).

object detection

Automatically generate a pdf report containing feature importance, baseline modelling, spurious correlation detection, and more, from a single command line input for any given ML CSV file

Like a CLIP + VQGAN. Except without a VQGAN.

[ICCV 2023] Code implementation for "Leaping Into Memories: Space-Time Deep Feature Synthesis"

OpenVision: A Fully-Open, Cost-Effective Family of Advanced Vision Encoders for Multimodal Learning. + A closer look in PyTorch!

Some Class Activation Map methods implemented in Pytorch for CNNs

Variational_Autoencode_Feature_Visualization

CLIP class/concept/facet/inner neuron visualization (CNN and ViT); plus: king-man+woman=queen in Vision Transformers

Master's thesis project: Analyzing the learnings of a 3D PointNet

A simple code for visualizing the features in networks, with samples for convenience.