Ashutosh Barkule

My Projects

• Indian Law LLM

  LLM to assist in the classification of crimes under the Indian Penal Code (IPC). The application processes crime descriptions and identifies relevant IPC sections, procedures, and recommendations.

• Canadian Accident Law LLM

  LLM to assist in providing legal advice based on Canadian accident laws for insurance purposes. The application processes descriptions of accident situations and identifies relevant legal sections, liability, procedures, and recommendations.

• ArcGIS Object Segmentation

  Specialist in enhancing ArcGIS with custom object detection capabilities for geospatial analysis. This role involved integrating advanced object detection functionalities into the ArcGIS environment, streamlining workflows and improving accuracy in spatial data interpretation.

• 3D Semantic Segmentation

  This project focuses on developing expertise in 3D semantic segmentation and object extraction using advanced machine learning techniques. The goal is to design a highly accurate 3D object detection model by leveraging Kernel Point Convolution (KPConv) capabilities within PyTorch, integrating data from LiDAR and 2D imagery, and ensuring rigorous data validation and post-processing.

• 3D Traffic Monitoring

  This project focuses on developing a state-of-the-art traffic monitoring system using monocular 3D object recognition techniques. The system leverages deep learning algorithms to accurately detect and monitor various road users, such as vehicles and pedestrians, using a single camera. This approach aims to provide a cost-effective and efficient solution for traffic monitoring in urban environments.

• 2D Computer Vision Projects

  This project showcases expertise in 2D computer vision, focusing on object detection and instance segmentation. It involves collaborative strategy, versatile model training, automated annotation, optimization, and advanced deployment techniques. Examples include Vehicle Damage Assessment, Road Assets Detection, Construction PPE Detection, and X-ray Scanning

• Real-Time Speech Translation

  This project focuses on real-time translation of spoken language using advanced speech recognition and language processing models. The application utilizes the Whisper model for transcribing speech in real-time, providing immediate translation and transcription.