nabarun goswami

Current: Specially Appointed Assistant Professor, University of Tokyo. Audio x AI researcher. Past: Project Lead at @nablas_inc, Tech Lead at Sony

[nabarun goswami @naba89]

Excited to share that I’m joining @MIL_UTokyo at The University of Tokyo as a Project Assistant Professor! 🎉 Working at the cutting edge of Speech × AI. 🇯🇵🔊🤖 #AI #SpeechTech

[nabarun goswami @naba89]

Honored to receive the Gold Reviewer Award and complimentary registration for @icmlconf 2026 🏅

Reviewing has been one of the most fulfilling parts of research for me, and I’m grateful for the opportunity to contribute to the community through constructive feedback.

#ICML2026

[東京大学 先端研 原田研究室 @MIL_UTokyo]

【研究室ブログ更新】 国際会議 Interspeech2025 の URGENT 2025 Challenge に参加したシステム "FUSE" について、論文の著者である Nabarun Goswami が解説記事を執筆しました！　ぜひご覧ください。
https://t.co/kFoofp1tjp

[東京大学 先端研 原田研究室 @MIL_UTokyo]

【研究室ブログ更新】 当研究室の Nabarun Goswami 特任助教（ @naba89 ）が、「Speech Enhancement （音声強調）」の国際的なコンペティションである The ICASSP 2026 URGENT Challenge (URGENT Challenge 2026) の Track 1 に参加し、第２位を獲得しました。
https://t.co/O1IVWgzeLt

[nabarun goswami @naba89]

@unilightwf If you do not have sudo access to apt install, you could install cmake via conda from conda-forge, provided you are on a conda environment.

[nabarun goswami @naba89]

@unilightwf Regarding not reading whole audio, for SE/MSS task, training could be done on fixed-length segments and for inference overlap-add could be used. That way we just need to read short segments from the file rather than the whole audio.

[nabarun goswami @naba89]

@unilightwf Speed: On my system, I get around 8-10 batches of 16 samples per second with 8 workers for online and around 15-16 batches per second for offline with same settings. But my model (~20M param) can only consume 2-3 batches per second, so online is not a bottleneck.

[nabarun goswami @naba89]

@unilightwf Additionally, it is faster to first read audio metadata and sample smaller segments, rather than reading the whole file and then chopping it up.

[nabarun goswami @naba89]

@unilightwf URGENT 2026 Track 1 baseline includes both on-the-fly and offline noise/distortion augmentation:
🔗 https://t.co/8lfUDqmdw1
Large models, using 8–16 DataLoader workers is typically enough to avoid I/O bottlenecks, even with sox/ffmpeg augmentations in on-the-fly mode.

[Roger K Moore @rogerkmoore]

If you missed my keynote at INTERSPEECH-2025 (or would like to see it again), it’s now available online at https://t.co/sjXWmsaz9L - my bit is Keynote 1 and it starts at 1:05:30

[nabarun goswami @naba89]

Great to be in Rotterdam for Interspeech 2025. 🇳🇱
#Interspeech2025

[nabarun goswami @naba89]

Traveling to Rotterdam 🇳🇱 for INTERSPEECH 2025!

I’ll be presenting our paper in the URGENT Challenge Special Session (Area 14, SS2) at 15:45.
Our system ranked 3rd in the challenge.

📄 https://t.co/EBai5fKDnR

#INTERSPEECH2025 #SpeechEnhancement #URGENTChallenge

[NABLAS株式会社 @nablas_inc]

【論文公開】
AI生成動画のフェイク検出に関する論文を公開しました。
https://t.co/Dw8G2aOHwZ

生成AIの発展で自然なフェイク動画が生成されるようになりました。これにより、従来の検出手法ではフレーム間の微細な不整合を捉えることが難しくなっています。
✅本研究では、オプティカルフローの残差を入力に追加する検出フレームワークを提案しました。通常の入力で各フレーム内のアーティファクトを検出し、もう一方では時間的な不整合を検出します。シンプルな解決策ながらすべての設定でベースラインモデルを上回る結果が得られました。

NABLASのフェイク検出「KeiganAI」はこちら
https://t.co/GlSHHn6kOl

[nabarun goswami @naba89]

@motokiomura Congratulations Omura san!!! 🥳🥳

[nabarun goswami @naba89]

🔗 My PhD work covers the following papers:

• HyperVQ: MLR-based Vector Quantization in Hyperbolic Space (TMLR)
👉 https://t.co/sCKxdA9UGx
• EDM-TTS: Efficient Dual-Stage Masked Modeling for Alignment-Free TTS (TMLR)
👉 https://t.co/3c2MdJSmc5

[nabarun goswami @naba89]

🎓 Successfully defended my PhD thesis today!
Title: Efficient Discrete Speech Modeling via Non-Autoregressive Methods for Joint Synthesis and Recognition
Grateful to my advisor, committee, and everyone who supported me on this journey. Onwards!🚀 #PhDDefense #SpeechAI #TTS #ASR

[nabarun goswami @naba89]

@motokiomura Congratulations!!

[nabarun goswami @naba89]

Excited to present my poster in ICLR!
Do drop by if you are around!

#50: T2V2: A Unified Non-Autoregressive Model for Speech Recognition and Synthesis via Multitask Learning

Poster Session 3
Friday, April 25, 2025
10:00 - 12:30
Hall 3 + 2B

#ICLR25 #ICLR