by Nimish Sanghi | Apr 16, 2021 | AI & ML, Nimish
1. Introduction NLP (Natural Language Processing) has popularised the use of self-attention based Transformers as a way to go for scalable NLP tasks. Transformers have almost replaced RNN (Recurrent Neural Network) based architectures for NLP. Recently, Deep Learning...
by Nimish Sanghi | Mar 15, 2021 | AI & ML, Nimish
In this post we will cover the concept of model interpretability. We will talk about its increasing importance as newer AI driven systems are getting adopted for critical and high impact scenarios. After the introduction, we will look into some popular approaches...
by Nimish Sanghi | Mar 8, 2021 | AI & ML, Nimish
In this post I will talk about the Masked Autoencoder for Distribution Estimation MADE which was covered in a paper in 2015 as linked above. I will follow the implementation from University of Berkeley’s Deep Unsupervised Learning course which can be found here....
by Nimish Sanghi | Feb 26, 2021 | AI & ML, Nimish
1. Introduction Reinforcement Learning (RL) is a sub topic under Machine Learning. It is one of the fastest growing disciplines helping make AI real. Combining Deep Learning with Reinforcement Learning has led to many significant advances that are increasingly getting...
by Nimish Sanghi | Jan 6, 2021 | AI & ML, Nimish
Text in natural scene images usually carries abundant semantic information. However, due to variations of text and complexity of background, detecting text in scene images becomes a critical and challenging task. This algorithm consists of a fully convolutional...
by Nimish Sanghi | Jan 6, 2021 | AI & ML, Nimish
Introduction Natural Language Processing requires sequential data processing and data at any particular time may/may not be related to data sent at earlier time interval. This makes NLP bit complex compare to computer vision problem as network needs to remember...
