In refining D4AA9:3ED787 (Narva) in 00h 06m 05s 719ms I have brought glory to the company. Praise Kier. 9⃣5⃣3⃣7⃣4⃣ 9⃣3⃣6⃣7⃣2⃣ 5⃣8⃣6⃣2⃣9⃣ 8⃣1⃣9⃣2⃣9⃣ 3⃣9⃣4⃣6⃣0⃣ #mdrlumon #severance lumon-industries.com

I'm an experienced audio programmer working with sound, speech, and music, and I’m looking to transition into speech recognition and voice interface engineering. I see this as the next frontier for user experiences, especially with the advent of real-time voice AI systems like GPT-4o voice and other low-latency, intelligent machine listening systems coming to market.

My background:

- I have 10+ years experience as a software engineer, most recently as Director of Engineering at a startup

- I have extensive experience in audio programming with a focus on sound classification, some experiments with speech, and lots more experience with music

- I'm self-taught in audio signal processing (completed a Coursera course)

- I've built sound classifiers using fast.ai in the past (e.g. environmental sounds, bird sounds)

    - https://github.com/aquietlife/whisp

    - https://github.com/aquietlife/flightsoffancy

Current situation:

- I'm engaged in a year-long, self-directed fellowship to dive deep into this field through self-study and mentorship

- I'm interested in exploring the origins of bias in speech recognition technology, particularly around marginalized voices (e.g. AAVE in the U.S.), disfluency, and low-resource languages

- I'm seeking advice on how to best use this year to prepare for meaningful work/research opportunities

- I'm interested in specializing in mitigating risks and biases in speech recognition systems

Goals:

- I want to focus on building an end-to-end speech recognition system (ideally with PyTorch)

- Ultimately I want to transition into speech recognition and voice interface systems work

- I'd like to avoid going back to school full-time

- I want to develop skills and knowledge to make an impact in this field

Questions:

1. What are the most crucial skills/knowledge areas to focus on?

2. Are there specific courses, books, or resources you'd recommend?

3. How can I build a portfolio that demonstrates competence in this field?

4. Are there particular companies or research labs that are good targets for someone with my background?

5. What are some common pitfalls to avoid in this transition?

6. What are some key challenges and approaches in mitigating bias in speech recognition systems? Are there specific resources or research groups focused on this aspect?

Any advice, personal experiences, or insights would be greatly appreciated. Thanks so much for anything you can share to help me along this journey :)

(edits: formatting)

Comments URL: https://news.ycombinator.com/item?id=41520413

Points: 1

# Comments: 2

http://www.metmuseum.org/blogs/digital-underground/2017/open...

The Metropolitan Museum of Art presents over 5,000 years of art from around the world for everyone to experience and enjoy. The Museum lives in three iconic sites in New York City—The Met Fifth Avenue, The Met Breuer, and The Met Cloisters. Millions of people also take part in The Met experience online.

Since it was founded in 1870, The Met has always aspired to be more than a treasury of rare and beautiful objects. Every day, art comes alive in the Museum's galleries and through its exhibitions and events, revealing both new ideas and unexpected connections across time and across cultures.

The Metropolitan Museum of Art provides select datasets of information on more than 420,000 artworks in its Collection for unrestricted commercial and noncommercial use. To the extent possible under law, The Metropolitan Museum of Art has waived all copyright and related or neighboring rights to this dataset using Creative Commons Zero. This work is published from: The United States Of America. You can also find the text of the CC Zero deed in the file LICENSE in this repository. These select datasets are now available for use in any media without permission or fee; they also include identifying data for artworks under copyright. The datasets support the search, use, and interaction with the Museum’s collection.

At this time, the datasets are available in CSV format, encoded in UTF-8. While UTF-8 is the standard for multilingual character encodings, it is not correctly interpreted by Excel on a Mac. Users of Excel on a Mac can convert the UTF-8 to UTF-16 so the file can be imported correctly.

Comments URL: https://news.ycombinator.com/item?id=13589891

Points: 9

# Comments: 2