Taesik Gong
@Taesik_MobileAI
Followers
370
Following
460
Media
10
Statuses
56
Assistant Professor at UNIST (CSE & AIGS) | Prev: @Cambridge_Uni @BellLabs @GoogleAI @MSFTResearch @kaistcsdept | Google PhD Fellow
Joined November 2016
Excited to share that our work on test-time adaptation with binary feedback has been accepted to #ICML2025!. Huge thanks to my amazing collaborators!.#ICML #ICML2025 #TTA.
We're happy to announce that our work on test-time adaptation with binary feedback has been accepted to #ICML2025!. We will discuss how we can utilize human feedback for TTA. Special thanks to collaborators.@sornswgn @JunsuKim97 @jinwoos0417 @Taesik_MobileAI @wewantsj !!
1
0
20
Thrilled to share that our paper “SelfReplay: Adapting Self-Supervised Sensory Models via Adaptive Meta-Task Replay” has been accepted to ACM #SenSys2025!. Grateful to collaborate with such an incredible team!🎉.
Pleased to share that our paper, “SelfReplay: Adapting Self-Supervised Sensory Models via Adaptive Meta-Task Replay,” is accepted to #SenSys2025!. Huge thanks to my collaborators, Jaehyun, Biniyam, @GaoleDai4, Prof. Mo Li, @Taesik_MobileAI, @kimin_le2, and @wewantsj! (1/2)
0
0
3
📄Join me for my #NeurIPS2024 poster presentation on 11th Dec! . "DEX: Data Channel Extension for Efficient CNN Inference on Tiny AI Accelerators". 📆 When: Wed 11 Dec 4:30 pm- 7:30 pm.📍 Where: East Exhibit Hall A-C 1405. #NeurIPS #NeurIPS24 #MachineLearning #AI #AIAccelerator
0
3
17
Excited to see our work on MLLMs with sensor data has been accepted at #EMNLP2024 (main, long)! Proud to be part of an incredible team. I look forward to continuing our journey in exploring how we can push the boundaries of AI with sensor data and MLLMs. 🎉.
🎉 Thrilled to share our paper “By My Eyes: Grounding Multimodal Large Language Models with Sensor Data via Visual Prompting” is accepted at #EMNLP2024 Main! 🙌 Huge thanks to Biniyam, Prof. @Taesik_MobileAI, @kimin_le2, and @wewantsj!. 📄 Preprint: (1/2)
0
0
19
📄 I am thrilled to share that our work from my time at Bell Labs has been accepted to NeurIPS 2024! Huge thanks to my amazing collaborators, @raswak and @ChulhongM ! 👏. Stay tuned for the camera-ready version!.#NeurIPS2024 @NeurIPSConf
0
2
38
📢 I am thrilled to start as an Assistant Professor at UNIST in the Department of CSE & AI Graduate School! . Huge thanks to Bell Labs, my amazing colleagues, and my advisor Prof. @wewantsj for the unconditional support throughout my journey.
8
2
107
RT @taeckyung_lee: Curious about *estimating test-time adaptation (TTA) accuracy* without labeled data or source data access? Check out our….
0
2
0
📢I'm thrilled to announce my new role as a visiting scholar at the @Cambridge_Uni, hosted by Prof. Cecilia Mascolo (@cecim)! Excited to collaborate with the world-class research groups and brilliant students here! 🌟
2
0
41
📄 Thrilled to announce that our research, "ATTA: Label-Free Accuracy Estimation for Test-Time Adaptation," has been accepted at #CVPR2024! Kudos to the awesome team: @taeckyung_lee, @sornswgn, and @wewantsj 👏. Stay tuned for our camera-ready version and code!
0
2
21
RT @raswak: 🔥 Only 5 days left to apply – so act fast! Let's turn this summer into an unforgettable one! It's all about breaking things for….
lnkd.in
This link will take you to a page that’s not on LinkedIn
0
6
0
RT @spdimitris: Join us for a research internship this summer in Cambridge! Topics: self-supervised learning, federated learning, parameter….
0
16
0
📄Join me for my #NeurIPS2023 poster presentation tomorrow!."SoTTA: Robust Test-Time Adaptation on Noisy Data Streams". 📆 When: Tue 12 Dec 10:45 am- 12:45 pm.📍 Where: Great Hall & Hall B1+B2 (level 1) # 1707. @NeurIPSConf. #NeurIPS #NeurIPS23 #MachineLearning #AI #TTA
0
3
18
📢 Looking for wonderful research internship opportunities? Join us at Nokia Bell Labs (Cambridge, UK)! We have multiple research internship positions available.
📢 Hello, future rockstars! It's that exciting time of year again when I invite you to join my Cambridge Lab for an unforgettable summer experience. If you're keen to make, shake, and break while engaging in curiosity-driven research, consider this your call. Reach out! 🤙🧑🔬🚀
0
4
27
0
0
0
TL;DR: Test-time adaptation usually assumes a stable testing environment, but in the real world, unexpected diversity exists in test data, such as new objects or noises. To combat this, we introduce SoTTA, a robust TTA algorithm for these realistic test streams.
1
0
0
📢 We've just released the arXiv preprint and code for our work "SoTTA: Robust Test-Time Adaptation on Noisy Data Streams" (NeurIPS '23):. 🖥️ Code: 📚 arXiv: #NeurIPS #NeurIPS23 #MachineLearning #AI #TTA.
github.com
This is the official PyTorch Implementation of "SoTTA: Robust Test-Time Adaptation on Noisy Data Streams (NeurIPS '23)" by Taesik Gong*, Yewon Kim*, Taeckyung Lee*, Sorn C...
1
0
12
Kudos to my amazing collaborators: @hai_yewon, @AdibaOrz , Yunxin Liu, @SungJuHwang1, @jinwoos0417, and @wewantsj.
0
0
0
🔍 Summary: We address the uncertainty of performance after domain adaptation. DAPPER estimates adapted performance with unlabeled data, leveraging the correlation between model outputs and model performance. DAPPER achieves a 93% similarity to ground-truth model accuracy.
1
0
0
🗣️Join me for my IMWUT (UbiComp) paper presentation tomorrow (Wed.) from 2 pm in Cozumel B!. 📄 Title: "DAPPER: Label-Free Performance Estimation after Personalization for Heterogeneous Mobile Sensing". #IMWUT #UbiComp #UbiComp2023
1
4
24
Excited to announce that our research on robust test-time adaptation for noisy streams has been accepted at #NeurIPS2023! Huge thanks to my incredible collaborators: @hai_yewon, @taeckyung_lee, @sornswgn, and @wewantsj. 🙌. Stay tuned for the code and camera-ready version!
0
3
33