Desh Raj

These are my notes from reading the Ultra-scale Playbook by HuggingFace.

A few days ago, I conducted a poll soliciting opinions about the following question: For the same parameter size, network architecture, and training data, which of the following models do you think would perform best at streaming ASR?

Book Link These are my notes taken while reading the RLHF book by Nathan Lambert.

Neural transducers are the most popular ASR modeling paradigm in both academia and industry.

This note is a brief introduction to approximation algorithms.

Since I am preparing for my GBO exam (which is a kind of qualifying exam where a committee evaluates your preparedness towards your PhD), my advisor Sanjeev Khudanpur suggested that in addition to preparing low-level details, I should also be able to answ…

In this series of notes we will review some basic concepts that are usually covered in an Intro to ML course.

In this series of notes we will review some basic concepts that are usually covered in an Intro to ML course.

In this series of notes we will review some basic concepts that are usually covered in an Intro to ML course.

In this series of notes we will review some basic concepts that are usually covered in an Intro to ML course.

In this series of notes we will review some basic concepts that are usually covered in an Intro to ML course.

In the previous post on the MVDR beamformer, we saw how speaker-specific “masks” can be used in conjunction with a multi-channel input signal to extract the speaker-specific signal in the presence of background noise or interfering speakers.

Guided source separation (GSS) is an unsupervised algorithm for target speech extraction, first proposed in the Paderborn submission to the CHiME-5 challenge.

In a previous note, we described the process of mask estimation using complex angular central GMMs that are used in guided source separation (GSS).

Speaker diarization is often formulated as a clustering of speaker embeddings.

In this note, I will review a popular clustering algorithm called spectral clustering.

In this note, we will describe how to estimate the parameters of GMM and HMM models using expectation-maximization method.

In this note, we will review the two most popular speaker embedding extraction methods, namely i-vectors and x-vectors.

Introduction Transformers are ubiquitous in deep learning today.

I attended the virtual ICASSP 2021, and this is a short post with my 3 key take-aways from the conference.