Segment Anything Model (SAM): a new AI model from Meta AI that can "cut out" any object, in any image, with a single click. The Segment Anything Model (SAM) produces high quality object masks from input prompts such as points or boxes, and it can be used to generate masks for all objects in an image. It has been trained on a dataset of 11 million images and 1.1 billion masks, and has strong zero-shot performance on a variety of segmentation tasks.

DeepSORT based Tracking is based on the following steps: Detection: Before objects can be tracked in each frame, they must be detected. This is done using a standard object detector like Faster R-CNN. Feature Extraction: Extract features from the detected objects. These features will help to match objects across frames. Data Association: Match detected objects with tracked objects from previous frames using both the bounding box overlap (using the Hungarian algorithm) and the feature similarity. Track Management: Update tracks or create/delete tracks as necessary.

YOLO8 is the latest version of the YOLO family, featuring enhanced capabilities such as instance segmentation, pose/keypoints estimation, and classification.

Detectron2 is an open-source modular computer vision framework built on top of PyTorch. It's developed by the Facebook AI Research (FAIR) team and serves as a successor to the original Detectron platform. Detectron2 is designed specifically for research and development purposes in the domain of object detection and segmentation. This is the official tutorial of detectron2. Here, we will go through some basics usage of detectron2, including the following: Run inference on images or videos, with an existing detectron2 model Train a detectron2 model on a new dataset

In this tutorial, we will be finetuning a pre-trained Mask R-CNN_ model on the Penn-Fudan Database for Pedestrian Detection and Segmentation_. It contains 170 images with 345 instances of pedestrians, and we will use it to illustrate how to use the new features in torchvision in order to train an object detection and instance segmentation model on a custom dataset.

CS224N: Natural Language Processing with Deep Learning. Session 1 is focussed on Word Vectors

Transformers from sratch (draft)

Demo of the Encodec paper proposes a model for audio compression, which takes an audio signal as input and produces a compressed representation of the signal.

Session 4 on CNN, ViT, GAN and Some practical concepts. Video and PDF attached with the published project.

This is part of the Deep learning course series by DEQUE AI. A simple example of a vanilla neural network on FashionMNIST dataset.

Session 3 on Transformers. This notebook demonstrates some standard Language tasks such as Sentiment Analysis, Text generation, Questions and Answer and so on.

In this example we implement a simple Deep Reinforcement Learning solution for the classic CartPole problem. We implement a custom reward function and then we use three different strategies to strike a balance between exploration and exploitation and use DQN to approximate Q-values: - Epsilon greedy exploration - Noisy Networks - Thompson Sampling

In this project, the Transformer architecture is applied to detect and pinpoint objects. The dataset used is Caltech 101 from the Caltech library. Following the research published in 2021 “An Image is Worth 16*16 Words”, this project will implement a similar architecture to detect objects. Using the self-attention mechanism on patches of images is a way to derive attention maps which help focus on what is relevant in the image.

This is an LSTM Seq2Seq model for Machine Translation. This is to compare earlier performance with Transformers.

Application for deep learning engineer role at Deque.ai.

Code for Developing Attention Model with Keras

Exploring transformer architecture that was introduced in the research paper "Attention is all you need."

Exploring "Attention is all you need" research paper (Transformer architecture) with machine translation from English to French.

RNN, LSTM, and Transformer comparisons

Audio Generation with AudioLDM

VToonify: Controllable High-Resolution Portrait Video Style Transfer

Keras attention exploration

Tweet data exploration

Keras Char RNN

A simple Keras Word RNN

A simple word to vector example

This is a image captioning task. I compared RNN and LSTM and handcrafted adding image features to LSTM in a different way.

This project solved a 6D pose estimation problem in computer vision. The pipeline is to first use a segmentation network (e.g., U-Net) to segment all objects in an image, and then use a learning-based method, ICP (iterative closest point)to estimate the pose for each segmented object

Toxic comment classification with tensorflow

Binary Classification of Happy and Sad images

Image classification analysis using Cifar10 database

A simple start project for learners to learn write the seq2seq model.

Brain tumor classification

This notebook is about performing an image classification task and examining the impact of data augmentation using a Generative Adversarial Network.

This notebook is about understanding and training a conditional Generative Adversarial Network for image generation.

This is a very simple, kind and easy-learning project for users who just start to learning computer vision or deep learning.

Classification for the MNIST dataset using CNN and Pytorch library.

Image classification model on the CIFAR10 dataset using Convolutional Neural Networks

Lung registration full functional job

Sentiment Analysis by using Tensorflow dataset Yelp-Restaurant-Review

Deformable lung image registration with a Dice score over 0.9 with only 10 Epochs

Data generation with using DCGAN

Various Machine Learning model Experiment with Tesnorflow Iris Dataset

Single click distributed training using pytorch

Single click distributed training using Tensorflow

Udacity - Deep Learning - Part 4 -Classifying Fashion-MNIST in pytorch

Udacity Deep Learning - Pytorch - Part 3 - Training Neural Networks in Pytorch

Udacity Deep Learning - Pytorch - Part 2 Neural Networks in Pytorch

Udacity Deep Learning Nano Degree - Pytorch - Part 1 - Intro to Tensors in pytorch

In this notebook we will train a scikit-learn and a cuML Random Forest Classification model. Then we save the cuML model for future use with Python's pickling mechanism and demonstrate how to re-load it for prediction. We also compare the results of the scikit-learn, non-pickled and pickled cuML models.

This Project does a custom object detection using YOLO model. The trained model was used to develop a tracking algorithm.

This book is a unique walk through for all aspirants that look forward to grasp the idea of supervised learning from scratch. This book shows the use of various metrics and hyper parameter tuning. Also this book sheds light into application of classifiers like KNN and so on... Happy learning....

This notebook is a basic operational procedure of statistical understanding of data and associated operations using numpy and pandas. Would be a great help for beginners to better understand the concepts in detail

The project is a simple elaboration of the concept of object detection from the tensorflow website.

This project represents a computer vision problem and solution with respect to activity recognition. The presented notebook is an example of a successfully compiled one that was able to identify activities from the predefined activities on recorded videos. The dataset used for the learning was UCF 50. All the methods including downloading and utilising the dataset is elaborated in the code base.

This Project does a custom object detection using YOLO model. The trained model was used to develop a tracking algorithm.

Built in MLFlow integration

Tensorflow MultiNode training

Tensorflow distributed training on multiple nodes

Built in Tensorboard Viewer

Convnext

Q&A with pytorch

Vtoonify in action