Utkarsh Madan

Skilled engineer specializing in computer vision, deep learning, Machine Learning Operations, and machine learning. Proficient in developing scalable and robust ML pipelines and deploying high-performance models for solving complex real-world problems. Committed to staying updated with technological advancements to continually improve the impact of work. Organized and dependable candidate successful at managing multiple priorities with a positive attitude. Willingness to take on added responsibilities to meet team goals.

Smart Web Scraping Project Using:- Python,Selenium ,Beautiful Soup Pandas, Numpy ,Visual Studio Code 

Objective- Made a Smart and Big Scrapper Project that First Extracts Details Every Past Valorant Tournament using Selenium and Beautiful Soup from site (www.liquipedia.gg).Then Extracted Stats of each Game in each Tournament from site(ww.vlr.gg).Then organized all data into csv format using pandas.Used Modular Programming approach in Python. 

GitHub Link:- https://github.com/UtkShowcase/AutomatedValorantScrapper_708Technologies 

Video Conversion and Frame Extraction Using Parallel Processing Using:- Python ,OpenCV ,Threading ,Multiprocessing ,FFmpeg , Pandas, Visual Studio Code 

Objective- Smart System that takes Videos as Input and can Convert Video ,Trim Video,Frames Extraction Using Parallel Computing at its Core (Threading,MultiProcessing).Used Object Oriented Concept Approach in Python. 

GitHub Link:- https://github.com/UtkShowcase/VideoDataUsageInterface 

Face Clustering and Detection Using:- TensorFlow ,RetinaFace ,FaceNet ,SkLearn ,SVC , Pandas, Numpy,Jupyter 

Objective- Collected Multiple photos of Multiple Individuals into a Single Dataset(Folder),and then Used Retina Face For Detecting Face ,then Converting Each Detected Face into an Embedding Using FaceNet,and then using SVC Unsupervised Algorithm to Cluster them into Groups.Hence Separating and Grouping Photos of all Individual photos into their Respective Output Folders. 

Object Detection and Tracking Using:- Python, PyTorch, Ultralytics , YoloV8 ,OpenCV, DeepSort, Jupyter 

Objective- Detecting Vehicle and Tracking With Using YoloV8 and DeepSort and label them with unique ID converting the Labeled Frames Into an Output Video.Used Object Oriented Concept To Make it Smart System. 

Object Detection and Training Using:- Python,MMDetection,OpenCV, Jupyter 

Objective- Trained Custom Detector on Aquarium Dataset Using MMDetection FrameWork 

Jupyter Notebook Link:- https://drive.google.com/le/d/1x0rVP6tCvsIP8CoXGEDM5h3tSo-4Maxq/view?usp=sharing