About Me
I am Tapabrata (also known by my nickname Rohan), a postdoctoral researcher in computer vision and machine learning applied to medical image analysis. I work at the Department of Engineering Science, University of Oxford, UK. My current research is on explainable AI (XAI) to make machine learning applications in biomedical imaging applications more trasparent and trustworthy.
Earlier I completed my PhD at the Department of Computer Science, University of Otago, NZ. My PhD research topic was on collaborative transfer learning of fine-grained visual data, with special focus on endemic endangered species of NZ. Before that, I completed BEng and MEng in Electrical Engineering.
I am originally from Calcutta, often referred to as the City of Joy and the cultural hub of India. From the busy city I moved to idyllic Dunedin, NZ for my doctoral studies about 3 years ago. Recently I have shifted to Oxford. In my spare time I like to go on long hikes in the countryside. I also enjoy playing pool occasionally.
Research Projects[Top]
 |
Current Project: Causal Inference enabled Computer Vision
Deep learning based vision systems have achieved near human performance in many applications in recent times. However, as these systems have become more powerful, they have also become more complex, and the way these AI driven systems make decision has become more opaque. This raises issues of trust of adopting these solutions freely in tasks of high consequence like medical applications. The need of the hour is therefore to make computer vision systems more explainable or trustable. This is my present work as part of the Seebibyte project at the University of Oxford.
|
 |
Previous Computer Vision Projects: 2012-2018
[1] 2015-18: Fine-grained Visual Categorization | Species Recognition [5 articles]
- My PhD thesis at Univ. of Otago, NZ (collab. with ICSI, Univ. of California, Berkeley.)
[2] 2014-15: Document Image Analysis | Optial Character Recognition [1 article]
- Joint project between CVPR Unit, Indian Statistical Institute and University de la Rochelle.
[3] 2013-14: Biomedical Image Analysis | Retinal Blood Vessel Segmentation [1 article]
- Joint project between IVPR Group, Jadavpur University and University of Muenster.
[4] 2012-13: Biometrics | Human Face and Facial Expression Recognition [3 articles]
- My MEng thesis and supported by University Grants Commission, Govt. of India.
|
Selected Publications [Top]
Journals
- R. Roy, T. Chakraborti and A. S. Chowdhury, "A deep learning-shape driven level set synergism for pulmonary nodule segmentation," Pattern Recognition Letters, Elsevier, vol. 123, pp. 31-38, 2019. [paper]
- T. Chakraborti, B. McCane, S. Mills, and U. Pal, "LOOP Descriptor: Local Optimal Oriented Pattern," IEEE Signal Processing Letters, vol. 25, no. 5, pp. 635-639, 2018. [paper]
- T. Chakraborti, D. K. Jha, A. S. Chowdhury, and X. Jiang, "A self-adaptive matched filter for retinal blood vessel detection," Machine Vision and Applications, Springer, vol.26, no.1, pp.55-68, 2015.
[paper]
- T. Chakraborti, K. Das Sharma, and A. Chatterjee, "A novel local extrema based gravitational search algorithm and its application in face recognition using one training image per class," Engineering Applications of Artificial Intelligence, Elsevier, vol.34, no.3, pp.13-22, 2014.
[paper]
Conferences
- T. Chakraborti, B. McCane, S. Mills, and U. Pal, "Fine-grained Collaborative K-Means Clustering", IVCNZ 2018 (best student paper award). [paper]
- T. Chakraborti, B. McCane, S. Mills, and U. Pal, "A Generalized Formulation for Collaborative Representation of Image Patches (GP-CRC)", BMVC 2017. [paper]
- T. Chakraborti, B. McCane, S. Mills, and U. Pal: "Collaborative representation based
fine-grained species recognition", IVCNZ 2016.
[paper]
- N. Tripathy, T. Chakraborti, M. Nasipuri, and U. Pal: "A scale and rotation invariant
scheme for multi-oriented Character Recognition", ICPR 2016.
[paper]
