Abhinav Dhere

Publications

• 5. Abhinav Dhere, Vikas Vazhayil and Jayanthi Sivaswamy, Fast detection of sulcal regions for classification of Alzheimer’s disease and Mild Cognitive Impairment. IEEE International Conference on Signal Processing and Communications, 2022


• 4. Abhinav Dhere and Jayanthi Sivaswamy, Explainable COVID detection using multi-scale attention in Chest X-Ray images. IEEE Journal of Biomedical & Health Informatics, 2022


• 3. Abhinav Dhere and Jayanthi Sivaswamy. Self-supervised learning for segmentation. arXiv:2101.05456, (2021).


• 2. Manish Sharma, Abhinav Dhere, Ram Bilas Pachori, and U Rajendra Acharya. An automatic detection of focal EEG signals using new class of time–frequency localized orthogonal wavelet filter banks. Knowledge-Based Systems, 118:217–227, 2017.


• 1. Manish Sharma, Abhinav Dhere, Ram Bilas Pachori, and Vikram M Gadre. Optimal duration-bandwidth localized antisymmetric biorthogonal wavelet filters. Signal Processing 34:87–99, 2017.

Projects

Fast detection of sulcal regions for classification of Alzheimer’s disease and Mild Cognitive Impairment

Sep 2021 - March 2022 | Research Project | Tools used: PyTorch
Explored methods based on height maps for classification of healthy, Alzheimer's disease and Mild Cognitive Impairment.

Explainable Covid Detection From Chest X-Ray Images

May 2020 - June 2021| Research Project | Tools used: PyTorch Proposed a novel architecture and loss function for COVID detection with clinically consistent explanations.

Self-supervised Learning For Kidney Segmentation

March - Oct 2019 | Research Project | Tools used: PyTorch
Proposed a novel proxy task for kidney segmentation using self-supervised learning. Demonstrated faster convergence & better performance compared to training from scratch.

SEGMENTATION OF BRAIN SULCAL REGIONS FROM 3D MESH

Feb - Oct 2018 | Research Project | Tools used: MATLAB, PyTorch
Developed a novel method for conversion of 3D mesh to an image representation. Used this image representation of human brain’s 3D structure for building a fast, non-iterative method to determine regions of sulcal folds.

3D SEGMENTATION IN HEART MRI

Jan - May 2018 | Course project for Medical Image Analysis | Tools used: PyTorch
Implemented a deep CNN for segmentation of myocardium & blood pool from heart MRI based on HVSMR dataset.

Face recognition and verification

Aug - Nov 2017 | Course project for Statistical Methods in AI | Tools used: Torch (Lua), OpenCV
Implemented a face recognition method and evaluated it on three independent datasets and tested the method on face recognition between twins using celebrity images.

Contrast Based Filtering for Salient Region Detection

Aug - Nov 2017 | Course project for Digital Image Processing | Tools used: MATLAB
Implemented a contrast based saliency detection algorithm. It decomposes the image into SLIC superpixels \& obtains a saliency measure from filters describing rarity and compactness.

Classification of Epileptic EEG signals

Feb - July 2016 | Undergraduate Project | Tools used: MATLAB
Classified EEG signals as focal vs non-focal with LS-SVM. Used a set of wavelet entropy features computed from proposed novel time-frequency localized orthogonal filter banks.

Experience

Education

Skills

Proficient: Python • MATLAB
Comfortable: C • C++
Familiar: Bash • Java

Proficient: PyTorch • LaTeX
Comfortable: Keras • Django