No description provided.

real-time fire detection in video imagery using a convolutional neural network (deep learning) - from our ICIP 2018 paper (Dunnings / Breckon) + ICMLA 2019 paper (Samarth / Bhowmik / Breckon)

Projects from the Deep Learning Specialization from deeplearning.ai provided by Coursera

Using DUCK-Net for polyp image segmentation. ( Nature Scientific Reports 2023 )

This deep learning application can detect Facial Keypoints (15 unique points). They mark important areas of the face - the eyes, corners of the mouth, the nose, etc.

Implementation of Logistic Regression, MLP, CNN, RNN & LSTM from scratch in python. Training of deep learning models for image classification, object detection, and sequence processing (including transformers implementation) in TensorFlow.

This repositary contain all my exercises and projects of Udacity Computer Vision Nanodegree Program

Google MediaPipe Javascript Demos (including live demos)

Genre Classification using Convolutional Neural Networks

Pre-trained VGG-Net Model for image classification using tensorflow

TensorFlow Lite object detection example for Raspberry Pi Zero

Python module for face recognition with OpenCV and Deep Learning.

Computer Vision Case Study in image recognition to classify an image to a binary class, based on Convolutional Neural Networks (CNN), with TensorFlow and Keras in Python, to identify from an image whether it is an image of a dog or cat. (Includes: Data, Case Study Paper, Code)

Segmenting WSIs using Deep Convolutional Neural Networks

Detect road anomalies such as cracks, potholes, and bumps using our trained YOLOv8 models with visual demo. Real-time detection via Streamlit and Flask app

For this project, we are going to detect rice leaf disease using CNN and serve the result via messenger chatbot. We will also implement this to an independent Android app.

Used Convolutional Deep Neural nets to extract features from the image

Resume Parsing app to extract information using AI

Violence recognition in streaming video using Transfer Learning and MoViNets. The project leverages state-of-the-art deep learning techniques to create an efficient and accurate violence detection system.

Some learnings and code for computer vision basics

We present the Automatic Helmet Detection System, a CNN model trained on image dataset that can detect motorbikes as well as riders wearing helmets.

The project aims at building a machine learning model that will be able to classify the various hand gestures used for fingerspelling in sign language. In this user independent model, classification machine learning algorithms are trained using a set of image data and testing is done. Various machine learning algorithms are applied on the datasets, including Convolutional Neural Network (CNN).

Semantic neural network to realize pixel-wise classification of 2D nano-material using Matlab

Applied YOLO model trained on COCO dataset to detect obstacles and Lane-Net model trained on tusimple.ai dataset for end-to-end lane detection. • Improved usability and response time by 50% using the combination and optimization of legacy codes of algorithms in assisted driving.

Convolutional Neural Network (CNN) was trained on 48x48 pixel grayscale images to predict 5 different emotions from images. Ten different models with different settings were trained to find the best model and The best model was able to predict 5 emotions from images with 88% training accuracy and 70% testing accuracy.

This is "ready from box" face recognition app, based on Mediapipe, dlib and face_recognition modules.

Implement SS-CNN "Dubey et al. Deep Learning the City ECCV16"

ConvNet (CNN) implementation to classify x-ray medical images

Project for detecting Autism and Dyslexia using ML. A POC for the KPMG Ideation Challenge, 2021.

Proposing a novel machine learning-based approach for real-time suspicious activity detection in surveillance videos to enhance public safety and prevent terrorism, theft, accidents, and criminal activities.