detectron2 backbone: resnet18, efficientnet, hrnet, mobilenet v2, resnest, bifpn

[EXPERIMENTAL] Demo of using PyTorch 1.0 inside an Android app. Test with your own deep neural network such as ResNet18/SqueezeNet/MobileNet v2 and a phone camera.

Classification and Gradient-based Localization of Chest Radiographs using PyTorch.

A simple TensorFlow 2 implementation of ResNet-18

Image classification done with Mindspore technology

This is a clean, opinionated and minimalist implementation of the training and evaluation protocols for a fancy new self supervised learning (SSL) method called maximum manifold capacity representations.

Classify Beatbox Genre using Mel Spectograms and Resnet18 Classifier

It is the image classification task to classify Diabetic-Retinopathy category using ResNet18, ResNet50 pretrained model. which is related to kaggle competition. The kaggle competition link can found below. https://www.kaggle.com/c/diabetic-retinopathy-detection#description

First capstone project I worked on as part of the 'Programming in Python for AI' nanodegree at Udacity.

Bird Species Classification Using Transfer Learning

Tumor detection and classification from abdominal ultrasound images using CenterNet with Contrastive Learning.

🔪 Elimination based Lightweight Neural Net with Pretrained Weights

Detect if a driver is drowsy or not using a face detector and eye classifiier.

Rice Species Classification using ResNet-18 and a Custom defined CNN, both using PyTorch. If you are getting started with PyTorch, then you may consider cloning this repo and start learning :)

Collision Avoidance Strategies with Jetracer Pro AI Kit.

AI based security system!

DeepLab v3+ and Floral-Net, a U-Net inspired CNN, for segmenting flowers and backgrounds in the Oxford Flower Dataset, enhancing accuracy through tailored architecture and class imbalance handling.

Repository for a Dockerized Streamlit application that predicts sketched digits through a ResNet18 classifier.

This repo contains deep learning projects for beginners.

Training neural nets with quantized weights on arbitrarily specified bit-depth