<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>AI | Sasan Jafarnejad | AI Researcher</title><link>https://sasan.jafarnejad.io/tag/ai/</link><atom:link href="https://sasan.jafarnejad.io/tag/ai/index.xml" rel="self" type="application/rss+xml"/><description>AI</description><generator>Hugo Blox Builder (https://hugoblox.com)</generator><language>en-us</language><lastBuildDate>Sat, 01 Jun 2024 12:00:00 +0100</lastBuildDate><item><title>PodFetch - A Podcast Digestion Pipeline! (WIP)</title><link>https://sasan.jafarnejad.io/project/podfetch/</link><pubDate>Sat, 01 Jun 2024 12:00:00 +0100</pubDate><guid>https://sasan.jafarnejad.io/project/podfetch/</guid><description>&lt;p&gt;I always wanted to be able to search in podcast transcripts and find the episodes that are relevant to my interests. It&amp;rsquo;s crazy still in 2024 Spotify is not able to give you &amp;ldquo;more like this&amp;rdquo; episodes, or &amp;ldquo;search in episode content&amp;rdquo; feature. So I used Whisper and started my own search engine for podcasts. It&amp;rsquo;s still in the early stages, and so far I&amp;rsquo;ve only re-written it once! The first version was using Solr for fulltext search. Now that LLMs are a thing, I would also like to add episode summarization feature to it. At the moment I&amp;rsquo;m adding speaker diarization to the pipeline, and the beauty of it is that the system is learning to recognize more and more speakers which reduces the need for manual labeling.&lt;/p&gt;</description></item><item><title>KITT - A Conversational Voice Assistant for Vehicles</title><link>https://sasan.jafarnejad.io/project/kitt-llm-voice-assistant/</link><pubDate>Fri, 31 May 2024 12:01:00 +0100</pubDate><guid>https://sasan.jafarnejad.io/project/kitt-llm-voice-assistant/</guid><description>&lt;p&gt;KITT is a conversational voice assistant for vehicles, capable of answering questions related to routing, finding points of interest, and is aware of the local weather and traffic conditions. The assistant, including a customizable speech synthesizer, is accessible through a graphical user interface that facilitates experimentation by simulating the change in time, origin, destination, and location of the car.&lt;/p&gt;
&lt;p&gt;KITT is open source and available on &lt;a href="https://github.com/sasan-j/kitt" target="_blank" rel="noopener"&gt;GitHub&lt;/a&gt; and was presented at the 2024 IEEE Vehicular Networking Conference (VNC).
You can try KITT on HuggingFace Spaces &lt;a href="https://huggingface.co/spaces/sasan/KITT" target="_blank" rel="noopener"&gt;here&lt;/a&gt;.&lt;/p&gt;</description></item></channel></rss>