GAMMA : Generative Augmentation for Attentive Marine Debris Detection
VAishnavi Khindkar, Janhavi Khindkar
Under Review
We propose an efficient and generative augmentation approach to solve the inadequacy concern of underwater debris data for visual detection. We use cycleGAN as a data augmentation technique to convert openly available, abundant data of terrestrial plastic to underwater-style images. Prior works just focus on augmenting or enhancing existing data, which moreover adds bias to the dataset. Compared to our technique, which devises variation, transforming additional in-air plastic data to the marine background.
An efficient and scalable architecture for underwater plastic detection and cleaning using Underwater Autonomous Vehicle (AUV) and CycleGans as Data Augmentation technique to convert in air plastic to underwater style.
Vaishnavi Khindkar, Janhavi Khindkar
Patented idea of solving the dataset problem of underwater object detection by using cyclegans on our own created dataset .
AUTONOMOUS UNDERWATER VEHICLE FOR PLASTIC DETECTION, PLASTIC PROCESSING AND CLEANING.
Janhavi Khindkar, Aishwarya Patki
Patented idea of solving underwater plastic pollution using AUV using object detection for detecting the plastics and classifying them into various categories.
Multiclass Image Classification for Aerial Vehicle on UCMerced Dataset using TSBTC. Published IEEE conference.
Janhavi Khindkar, Sudeep Thepade
This project aims at muticlass classification of remote sensing image dataset.The model developed for classification is a fusion model of spatial features with dct features.3-layer fusion model of cnn is used with dct and lbp to improce the accuracy of prediction.It uses TBSTC as image processing unit for generating spatial and temporal features. The proposed model is the current SOTA.