Ameya Joshi

Ameya Joshi

I am a Research Scientist at InstaDeep AI, Boston, interested in multi-modal and sequence models. My current research focuses on building models for genomics and proteomics.

Sitemap

A list of all the posts and pages found on the site. For you robots out there is an XML version available for digesting as well.

Pages

Posts

Future Blog Post

less than 1 minute read

Published:

This post will show up by default. To disable scheduling of future posts, edit config.yml and set future: false. Read more

Kernels Part 1: What is an RBF Kernel? Really? Permalink

less than 1 minute read

Published:

An intriguing article. To look at an RBF kernel as a low pass filter is something novel. It also basically shows why RBF kernels work brilliantly on high dimensional images. Given that your image features generally lie in a continuous domain, an RBF kernel generally can fit smooth solutions and thereby create more relevant separating hyperplanes,especially in case of multiple classes. Read more

An Interesting History of Computer Vision Permalink

less than 1 minute read

Published:

Dr. Fei Fei Li from Stanford discusses the advent and growth of computer vision in recent years. Particularly intersting is her recent research on multimodal interactions and large scale visual recognition. This has been primarily made possible due to the growth in GPU technology. I hope to try out Theano and Caffe for deep learning in this scenario soon. Read more

VLAD- An extension of Bag of Words

4 minute read

Published:

Recently, I was a participant at TagMe- an image categorization competition conducted by Microsoft and Indian Institute of Science, Bangalore. The problem statement was to classify a set of given images into five classes: faces, shoes, flowers, buildings and vehicles. As it goes, it is not a trivial problem to solve. So, I decided to attempt my existing bag-of-words algorithm on that. It worked to an extent, I got an accuracy of 86% approximately with SIFT features and an RBF SVM for classification. In order to improve my score though, I decided to look at better methods of feature quantization. I had been looking at VLAD (Vector of Locally Aggregated Descriptors): A first order extension to BoW for my Leaf Recognition project. Read more

The Traveling Salesman Problem - Using Ant Algorithms

12 minute read

Published:

As I had promised, this post will be about using the Ant algorithms I had discussed in the previous post to solve a complex computational problem. But, before we go on, let us have a look again at Ant Colony optimization. Read more

Color Predicates

2 minute read

Published:

It’s been a long time since I posted something here. But here it is. Read more

The Start of An Age

less than 1 minute read

Published:

It has been several years since blogging took the world by storm. And it has progressed from being a way to just put down your thoughts for the world to read to being an entirely new way to view the world. Read more

portfolio

publications

talks

teaching

Teaching experience 1

Undergraduate course, University 1, Department, 2014

This is a description of a teaching experience. You can use markdown like any other post. Read more

Teaching experience 2

Workshop, University 1, Department, 2015

This is a description of a teaching experience. You can use markdown like any other post. Read more

uploads