Suyash P. Awate Suyash P. Awate
Associate Professor
Computer Science and Engineering Department, Indian Institute of Technology (IIT) Bombay
Office: A-214, Kanwal Rekhi Building
Email: my-first-name @cse.iitb.ac.in
   
Research Publications Teaching Students CV Personal
 
Image Segmentation and Classification with
Kernel Methods and Deep Neural Networks
 
Jena R, Awate SP
A Bayesian neural net to segment images with uncertainty estimates and good calibration
Information Processing in Medical Imaging (IPMI) 2019, xx-xx, Springer LNCS xx
(podium presentation, acceptance rate ~15%)
 
Shah M, Merchant SN, Awate SP
MS-Net: Mixed-supervision fully-convolutional networks for full-resolution segmentation
Medical Image Computing and Computer Assisted Intervention (MICCAI) 2018; 21(4):379-387
Springer LNCS 11073
(in top 10 papers nominated for MICCAI Young Scientist Award, podium presentation, acceptance rate 5%)
 
Shah M, Merchant SN, Awate SP
Abnormality detection using deep neural networks with robust autoencoding and semi-supervision
IEEE Int. Symposium on Biomedical Imaging (ISBI) 2018
(podium presentation, acceptance rate 17%)
 
Kumar N, Rajwade A, Chandran S, Awate SP
Kernel generalized Gaussian and robust statistical learning for abnormality detection in medical images
IEEE Int. Conf. Image Processing (ICIP) 2017, 4157-4161
(among top 10 finalists for Best Paper / Best Student Paper Award from 3000+ submissions,
student travel award)
 
Kumar N, Rajwade A, Chandran S, Awate SP
Kernel generalized-Gaussian mixture model for robust abnormality detection
Medical Image Computing and Computer Assisted Intervention (MICCAI) 2017; 20(3):21-29
Springer LNCS 10435
 
Shah M, Singha S,Awate SP
Leaf classification using marginalized shape context and shape+texture dual-path deep convolutional neural network
IEEE Int. Conf. Image Processing (ICIP) 2017, xx:xx-xx
(podium presentation)

Our Dual-Path CNN. “BN + ReLU” batch normalization followed by ReLU activation; N number of classes.


Example Leaves Differing in Shape and Texture. Leaf images input to the CNN path (left path in Figure 1) primarily capturing shape.


Marginalized Shape Context for 3 leaves: the leaves on the left and the middle belong to the same species.
 
Related Works