Ok, so I don't really maintain a blog. But I am active on Quora, the question and answering website. So I thought I'll try posting atleast those answers of mine, which happen to reach a blog-sized post on Quora. I am currently the most viewed writer in Computer Vision, and one of the top ten in Convolutional Neural Networks, Artificial Neural Networks, and Machine Learning. Quora made me one of their Top Writers 2016 and recently again for 2017. Last year, I got this neat bag as well as the chance to join their annual Top Writer meetup - which was an amazing experience! This year I got a really cool jacket from Quora - will post a selfie soon! I should mention that everything I write represents my personal views, and not the views of my present or past employers.
But before that, I wrote a little report on a major workshop Prof. Andrew Davison and myself organized in London which is available here. The title of the workshop was "Realtime 3D scene understanding in year 2020: The Role of computation in computer vision", and we managed to get a number of very high profile speakers.
Here is a proposal I wrote for the Onsager Fellowship in Robot Vision at NTNU in Trondheim, Norway. It got accepted and I was offered a tenure-track faculty position with funding for 2 PhD students. But I ended up moving to my current job instead. It showcases what I would have loved most to work on if real-life hadn't happened! :-)
I was also Associate Editor of Book Reviews for the International Association of Pattern Recognition (IAPR) quarterly newsletter from January 2013 through December 2016. My introduction in the newsletter can be found here. Besides coordinating book reviews written by other researchers, I have myself written reviews for two books:
1. Automatic Calibration and Reconstruction for Active Vision Systems, by Beiwei Zhang and Y. F. Li. (review - direct link)
2. Image Registration: Principles, Tools and Methods, by A. Ardeshir Goshtasby. (review - inside newsletter Pg 25)
My Quora profile is located here. Here are my answers to:
How was the ICLR 2018 conference?
What are the most interesting CVPR 2017 papers and why? I wrote this on my way back from Honolulu.
What are the most interesting CVPR 2016 papers and why? This is really a summary of my fresh thoughts right at the end of CVPR 2016.
How do I build a self driving car from scratch? This answer actually got featured on Quora's twitter feed as well, here.
Are the majority of machine learning careers simply using high-level libraries?
Is a PhD from a top university worth it if I am already doing deep learning research at a startup?
What is the next big thing in AI after deep learning and adversarial machine learning?
Everyone seems to want to study machine learning or be a data scientist. Are we heading for a talent bubble?
Is AI over-hyped in 2017? Also published by Quora Twitter feed and IB Times.
How does one prepare for a computer vision research scientist interview?
Can neural networks create consciousness?
What is a typical day like working on self driving cars as a full-time job? This also got published in Forbes Magazine, here.
Is deep learning just overfitting? Some random thoughts on the question.
Why does deep learning only work well on unstructured data?
What are some current research areas in Simultaneous Localization and Mapping (SLAM)?
Will the machine learning bubble burst in 2017?
Why is xgboost given so much less attention than deep learning despite its ubiquity in winning Kaggle solutions?
What will be the most important social change that augmented reality will bring?
What is an intuitive explanation of Deep Residual Networks?
Would getting into the field of Machine Learning until 2023 be too late or a bad idea?
What books have you read in 2016?