This repository consists of all the code required for similar 2-D pose detection in dance videos. This can used for any type of pose estimation application to find the similarity.

Implementation of the paper "Speech emotion recognition with deep convolutional neural networks" by Dias Issa Et al.

The project is a concoction of research (audio signal processing, keyword spotting, ASR), development (audio data processing, deep neural network training, evaluation) and deployment (building model artifacts, web app development, docker, cloud PaaS) by integrating CI/CD pipelines with automated tests and releases.

This is a web application that takes different kind of inputs(real-time, image, video) from the user and display the emotion based on the facial expressions.

Deep convolutional autoencoder for image denoising

An attempt to predict the Stock Market Price using Long Short Term memory and plot its chart. By tweaking different hyper parameters, we get different trained models. The aim of this project is to identify the relation hidden in these hyper parameters.

AI decoding of cvbs video

Deep Learning based wall/corridor following P3AT robot (ROS, Tensorflow 2.0)

This is a web application which can be used for data annotation for tasks like face detection, facial emotion classification, facial emotion detection

it able to detect ten type USA sign language. which is Okay, Peace, Thumbs up, Thumbs down, Call me, Stop, I Love You, Hello, No, Smile.

Image classification on famous DogVsCat dataset

A Neural Network playing World of Warcraft

Understanding Machine Learning and Unleash the Power of TensorFlow by Training Classification models for Image Recognition.

The Tensorflow version of EfficientSAM

Siamese Model for DeepFace Verification(similar to FaceNet architecture) with TensorFlow Serving

A product recommendation for E-commerce based on image similarity.

安师大教务系统验证码自动识别

Bank Transaction Data Generator