index.xml

<?xml version="1.0" encoding="utf-8" standalone="yes" ?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>My Portfolio Website</title>
    <link>/</link>
    <description>Recent content on My Portfolio Website</description>
    <generator>Hugo -- gohugo.io</generator>
    <language>en-us</language>
    <lastBuildDate>Sun, 15 Oct 2017 03:29:08 -0700</lastBuildDate>
    
	<atom:link href="/index.xml" rel="self" type="application/rss+xml" />
    
    
    <item>
      <title>Cope | Alexis Collado</title>
      <link>/cope/</link>
      <pubDate>Sun, 15 Oct 2017 03:29:08 -0700</pubDate>
      
      <guid>/cope/</guid>
      <description>The Challenge A startup called Cope hired me to create a minimum viable product for their new idea — tracking mental health. It was the first project where I handled mobile app design, and I was very excited to learn the intricacies of the iOS platform.
My clients John and Kat have done some preliminary interviews with psychologists and psychiatrists to get their side of the picture. They have established a user persona, a business model canvas, and several startup prep work for the product to take off.</description>
    </item>
    
    <item>
      <title>360 Pro | Alexis Collado</title>
      <link>/360pro/</link>
      <pubDate>Sun, 15 Oct 2017 03:29:08 -0700</pubDate>
      
      <guid>/360pro/</guid>
      <description>The Challenge Meeting with Chappy Callanta of 360 Pro led us to discover the root problem — the implementation of their performance testing framework. The gym uses a Numbers document on one iPad to record all tests, and relies on inefficient manual procedures to do the work. The document calculates a total pro score based on an athlete&amp;rsquo;s performance for specific exercises, but the coach using it has to refer to several norms before assigning a pro score.</description>
    </item>
    
    <item>
      <title>TEDxADMU | Alexis Collado</title>
      <link>/tedx/</link>
      <pubDate>Sun, 15 Oct 2017 03:29:08 -0700</pubDate>
      
      <guid>/tedx/</guid>
      <description>Context The Ateneo Junior Marketing Association always outsourced their website development to different organizations in the university. This year, they chose User Experience Society, the organization I founded to create a strong digital presence for their flagship event. TEDxADMU is acclaimed among the youth and is marked by the many up and coming trailblazers in the Philippine community.
The Challenge TEDxADMU always faced the same problems every year. How can we elevate the number of applications we receive?</description>
    </item>
    
    <item>
      <title></title>
      <link>/news/</link>
      <pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate>
      
      <guid>/news/</guid>
      <description>August 2025: Our paper Koel-TTS: Enhancing LLM based Speech Generation with Preference Alignment and Classifier Free Guidance got accepted at EMNLP 2025! August 2025: Our papers HiFiTTS-2 and NanoCodec got accepted at Interspeech 2025! July 2024: Our paper Improving Robustness of LLM-based Speech Synthesis Models by Learning Monotonic Alignment got accepted at Interspeech 2024! June 2024: I received the William S.C. Chang Best Ph.D. Dissertation Award in Electrical and Computer Engineering (ECE) Department at UC San Diego!</description>
    </item>
    
    <item>
      <title></title>
      <link>/preprints/</link>
      <pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate>
      
      <guid>/preprints/</guid>
      <description> Adversarial Deepfakes: Evaluating Vulnerability of Deepfake Detectors to Adversarial Examples Shehzeen Hussain*, Paarth Neekhara*, Malhar Jere, Farinaz Koushanfar, Julian McAuley
[ * Equal Contribution ]
[ paper, video examples ]  </description>
    </item>
    
    <item>
      <title></title>
      <link>/publications/</link>
      <pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate>
      
      <guid>/publications/</guid>
      <description>Koel-TTS: Enhancing LLM based Speech Generation with Preference Alignment and Classifier Free Guidance Shehzeen Hussain, Paarth Neekhara, Xuesong Yang, Edresson Casanova, Subhankar Ghosh, Mikyas T Desta, Roy Fejgin, Rafael Valle, Jason Li
Conference on Empirical Methods in Natural Language Processing EMNLP 2025, ICML 2025 Workshop on Machine Learning for Audio AI Heard That!
[ paper, ICML Workshop, audio examples ]
 NanoCodec: Towards High-Quality Ultra Fast Speech LLM Inference Edresson Casanova, Paarth Neekhara, Ryan Langman, Shehzeen Hussain, Subhankar Ghosh, Xuesong Yang, Ante Jukić, Jason Li, Boris Ginsburg</description>
    </item>
    
    <item>
      <title></title>
      <link>/research_experience/</link>
      <pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate>
      
      <guid>/research_experience/</guid>
      <description>NVIDIA Corporation July 2023 - Present
Santa Clara, California
Senior Research Scientist - Conversational AI
Investigating Large Language Models for multi-lingual voice conversion and robust text-to-speech synthesis.
 NVIDIA Corporation June 2022 - September 2022
Santa Clara, California
Research Intern - Deep Learning
Designed deep neural networks for multi-lingual voice conversion and text-to-speech synthesis. Developed state-of-the-art zero-shot voice conversion framework with speaker-adaptive pitch and duration controllability [ paper ]</description>
    </item>
    
  </channel>
</rss>