1 code implementation • 1 Apr 2024 • Taeckyung Lee, Sorn Chottananurak, Taesik Gong, Sung-Ju Lee
We propose the prediction disagreement as the accuracy estimate, calculated by comparing the target model prediction with dropout inferences.
no code implementations • 11 Dec 2023 • Taesik Gong, Si Young Jang, Utku Günay Acer, Fahim Kawsar, Chulhong Min
The advent of tiny AI accelerators opens opportunities for deep neural network deployment at the extreme edge, offering reduced latency, lower power cost, and improved privacy in on-device ML inference.
1 code implementation • NeurIPS 2023 • Taesik Gong, Yewon Kim, Taeckyung Lee, Sorn Chottananurak, Sung-Ju Lee
To address this problem, we present Screening-out Test-Time Adaptation (SoTTA), a novel TTA algorithm that is robust to noisy samples.
no code implementations • 7 Sep 2023 • Taesik Gong, Josh Belanich, Krishna Somandepalli, Arsha Nagrani, Brian Eoff, Brendan Jou
Speech emotion recognition (SER) models typically rely on costly human-labeled data for training, making scaling methods to large speech datasets and nuanced emotion taxonomies difficult.
no code implementations • 2 Sep 2022 • Hyungjun Yoon, Hyeongheon Cha, Hoang C. Nguyen, Taesik Gong, Sung-Ju Lee
Our evaluation with four different IMU sensing tasks shows that IMG2IMU outperforms the baselines pre-trained on sensor data by an average of 9. 6%p F1-score, illustrating that vision knowledge can be usefully incorporated into IMU sensing applications where only limited training data is available.
1 code implementation • 10 Aug 2022 • Taesik Gong, Jongheon Jeong, Taewon Kim, Yewon Kim, Jinwoo Shin, Sung-Ju Lee
Test-time adaptation (TTA) is an emerging paradigm that addresses distributional shifts between training and testing phases without additional data acquisition or labeling cost; only unlabeled test data streams are used for continual model adaptation.
no code implementations • 22 Nov 2021 • Taesik Gong, Yewon Kim, Adiba Orzikulova, Yunxin Liu, Sung Ju Hwang, Jinwoo Shin, Sung-Ju Lee
However, various factors such as different users, devices, and environments impact the performance of such applications, thus making the domain shift (i. e., distributional shift between the training domain and the target domain) a critical issue in mobile sensing.