Heng-Jui Chang

🎓 PhD Candidate @MIT_CSAIL

[Heng-Jui Chang @hjchang87]

📣🎉 Excited to announce that our paper was accepted to #INTERSPEECH2023!
https://t.co/vEOHdv2maV
https://t.co/3dFTD1DTi3

💡 Speaker-invariant Clustering (Spin)
1️⃣ Disentangles speaker
2️⃣ Preserves content
3️⃣ Benefits speech recognition & acoustic unit discovery

[Heng-Jui Chang @hjchang87]

@dearu_bubble #NMIXXbubbleFONT
1. 규진
2. 배이
3. 설윤
4. 릴리
5. 지우
6. 해원

[AI at Meta @AIatMeta]

We’re open-sourcing Perception Encoder Audiovisual (PE-AV), the technical engine that helps drive SAM Audio’s state-of-the-art audio separation.

Built on our Perception Encoder model from earlier this year, PE-AV integrates audio with visual perception, achieving state-of-the-art results across a wide range of audio and video benchmarks. Its native multimodal support can assist people in everyday tasks, including sound detection and richer audio-visual scene understanding.

🔗 Read the paper: https://t.co/RLWJOgG2uz
🔗 Download the code: https://t.co/1L5ZqCZlxq

[Yung-Sung Chuang @YungSungChuang]

Scaling CLIP on English-only data is outdated now…

🌍We built CLIP data curation pipeline for 300+ languages
🇬🇧We train MetaCLIP 2 without compromising English-task performance (it actually improves!
🥳It’s time to drop the language filter!

📝https://t.co/pQuwzH053M

[1/5]

🧵

[Heng-Jui Chang @hjchang87]

💡Bridging speech, sound, & music representations with one universal model?

We introduce USAD ✅
📚 Distills knowledge from domain-specific SSL models
🎯 Matches expert models across speech/audio/music tasks

📄 https://t.co/IHkZIGbTZg
🧑‍💻 https://t.co/TQ0V4wv7bA

[Yung-Sung Chuang @YungSungChuang]

(1/5)🚨LLMs can now self-improve to generate better citations✅

📝We design automatic rewards to assess citation quality
🤖Enable BoN/SimPO w/o external supervision
📈Perform close to “Claude Citations” API w/ only 8B model

📄https://t.co/FHj54HiC6i
🧑‍💻https://t.co/nQa87KkYMo

[Alexander H. Liu @alex_h_liu]

Presenting 2 works at #ICLR tomorrow!

📃Generative Pre-training for Speech with Flow Matching
📍5/9 (Wed) Hall B #68, 10:45am-12:45pm

📃Listen, Think, and Understand
📍5/9 (Wed) Hall B #60, 4:30pm-6:30pm

Please stop by if you're interested! More details...👇

[Hongyin Luo @lhyTHU]

(1/4) 💡Natural language embedded program (NLEP) is all you need for symbolic AND natural language tasks.

🚀NLEP outperforms ChatGPT-4, CoT, & PoT/PAL, without any task-specific example.

🎢NLEP makes small LMs outperform GPT-3 without fine-tuning!

https://t.co/3fLVlvUwtf

[Yung-Sung Chuang @YungSungChuang]

(1/5)🚨Can LLMs be more factual without retrieval or finetuning?🤔 -yes✅

🦙We find factual knowledge often lies in higher layers of LLaMA
💪Contrast high/low layers can amplify factuality & boost TruthfulQA by 12-17%

📝https://t.co/HyyIV782Tl
🧑‍💻https://t.co/KI56Gk3zRT

#NLProc

[Hongyin Luo @lhyTHU]

Excited to present our research at #acl2023 ! We found that self-trained entailment models with 350M parameters can outperform strong few-shot large language models with more than 100B parameters on several language understanding tasks. (1/4) https://t.co/DtVQt5QUcj

[Yuan Gong @YGongND]

(1/2) Introduce the IS23 Whisper-AT paper. We usually believe noise-robust ASR models' representations are noise-invirant. But we show a surprising finding that while Whisper is very robust against real-world background sounds, its representation is actually NOT noise-invariant.

[Yung-Sung Chuang @YungSungChuang]

Can language models help us do better search?🤔

🎉In #ACL2023 findings, we present EAR pipeline:

🎲sample multiple queries from LM
🎯rescore to select best query
🔍BM25 search
💪boost OpenQA accuracy to beat DPR/GAR

arxiv: https://t.co/7qdUXp31Q0
code: https://t.co/nxnDbZ6FXx

[Yung-Sung Chuang @YungSungChuang]

🚨We release SAIL-7B⛵️️a search-augmented instruction-tuned LM with:

🦆Real-time connecting to DuckDuckGo
🔍Explicitly filter out distracting search results
👨‍🏫Instruction following

Outperforms ChatGPT and Vicuna!🦙

demo: https://t.co/d8SqaAhNYf
arxiv: https://t.co/sQ4EbOUKpQ

[Andrew Rouditchenko 🇺🇦 @arouditchenko]

🗣️ Whisper is great for speech recognition, but it only recognizes ~100 languages. What if it wasn't trained on the language that you speak?

Happy to introduce my #INTERSPEECH2023 paper comparing Whisper and XLS-R for adaption to unseen languages!

https://t.co/GBW3rQ8L4N

[Heng-Jui Chang @hjchang87]

@ParcolletT results can be found on the cmt website!

[Yung-Sung Chuang @YungSungChuang]

📢New Paper Alert!!🚀
https://t.co/NaRoTRXB12

Does ChatGPT have the ability to check facts by itself?🤔
We designed a simple, few-shot, unified chain-of-thought prompting pipeline that can do:
🔹Fact-checking ✅
🔹Stereotype detection 🚫
🔹Hate speech detection 🙅 (1/2)

[Mirco Ravanelli @mirco_ravanelli]

Exciting news! 2 open positions for #PhD students to join our team and work on cutting-edge #research in #deeplearning, #conversationalAI, #speech tech, & sequence processing. Check out more details here and apply as soon as possible if interested:

https://t.co/CbwC7hM0DN

[Jia-Bin Huang @jbhuang0604]

It's time to let go

[Hung-yi Lee (李宏毅) @HungyiLee2]

Workshop on Self-supervised Learning for Audio and Speech Processing @ AAAI 2022 starts at 8:50 a.m., EST (9:50 p.m. GMT+8), February 28. If you want to hear about exciting new advances in self-supervised learning, don't miss it.
https://t.co/RBcP8jq5yV

[Heng-Jui Chang @hjchang87]

Excited to announce our paper was accepted to @ieeeICASSP 2022!

[Heng-Jui Chang @hjchang87]

Our DistilHuBERT model is released!
Thanks to @PatrickPlaten @leo19941227 @HungyiLee2 @ntu_spml!

paper: https://t.co/57gQTHaEa8
pre-training and inference code: https://t.co/mGHZlgkXsm