Chethana Badagi

Dimaag-AI

02.2022 - Current

Automated Pruning System for Vineyards.
Built an automated robotic pruning system to provide an efficient, consistent and cost-effective solution for vineyard owners.
Disparity estimation for stereo images using SGBM and RAFT.
Generation and optimization of point clouds (Open3d).
Getting segmentation for a point cloud from an image segmentation.
Extracting useful information from vine tree images using OpenCV.
Analyse data and develop different ML models and Graph Neural Networks for classification of a good or bad cane.
Graph Neural Networks for classification of good or bad cane.
Data creation, cleaning, EDA and pre processing.
Object detection using YOLOv5, Paddle Detection.
Image segmentation using SOLO, Mask2Former.
Development of AI based pruning based on user interactions.
Unet for semantic segmentation.
Feature extraction, analysis and testing from Images using OpenCV and 3D Point Clouds.
Video Productivity Analysis.
Training Few Shot Object Detection based on mmfewshot for custom dataset.
Video Label Assist using ms-tc action segmentation models, temporal pooling.
Binary Search, ASFormer, different versions of SAM(Segment Anything Model) and SAM2 video predictor.
Video Feature extraction using SlowFast.
Action Recognition using SlowFast.
Activity sequence analysis for the results obtained from Slowfast model.
Python Unit testing using Pytest.