AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Beyond Hearing: Learning Task-agnostic ExG Representations from Earphones via Physiology-informed Tokenization

Hyungjun Yoon, Seungjoo Lee, Yu Wu, XiaoMeng Chen, Taiting Lu, Freddy Yifei Liu, Taeckyung Lee, Hyeongheon Cha, Haochen Zhao, Gaoteng Zhao, Dongyao Chen, Cecilia Mascolo and 2 more

ICLR 2026 · Label-efficient and robust learning › Recognition and long-tail · approach: Dense features and foundation priors · 1 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add free-living pre-trainingPiMT · DailySenseavg F1: 81.9% to 87.6%Improved
add free-living pre-trainingPatchTST · DailySenseavg F1: 72.4% to 73.2%Improved
add pre-training to Bidirectional-Mamba without PiMTBidirectional-Mamba · DailySenseavg F1: 0.779 to 0.813Improved
add pre-training to Bidirectional-Mamba without PiMTBidirectional-Mamba · DailySensegaze angular error: 6.53 to 6.23Improved
add pre-training to Transformer without PiMTTransformer · DailySenseavg F1: 0.724 to 0.732No clear effect
use Mamba backbone instead of Transformer (ablation)PiMT · DailySenseF1: 6.4% improvement without pre-training and 8.5% with pre-trainingImproved
use 12-band filter bank instead of 1/2/4 bands (ablation)PiMT tokenization · DailySenseavg F1: average 4.6% F1-score gainImproved
add autoencoding pre-training task (ablation)PiMT · DailySenseavg F1: 0.873 -> 0.876Improved
add masked reconstruction pre-training task (ablation)PiMT · DailySensegaze angular error: 6.10 -> 6.00Improved
add amplitude reconstruction pre-training task (ablation)PiMT · DailySensegaze angular error: 6.19 -> 6.00Improved
add phase reconstruction pre-training task (ablation)PiMT · DailySensegaze angular error: 5.98 -> 6.00No clear effect
use PiMT tokenization instead of PatchTSTPiMT · DailySense (no pre-training)avg F1: 0.724 -> 0.819Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy