ProtoSSL discovers generalizable prototypes from unlabeled time-series via self-supervision and assigns them to new tasks for interpretable predictions, outperforming supervised baselines in low-data regimes on ECG datasets.
Title resolution pending
3 Pith papers cite this work, alongside 637 external citations. Polarity classification is still indexing.
3
Pith papers citing it
637
external citations · OpenAlex
representative citing papers
A framework and model detect mismatches between text and voice emotions in journaling using a controlled TTS dataset and asymmetric attention architecture, achieving macro-F1 of 0.711.
Moshi is the first real-time full-duplex spoken large language model that casts dialogue as speech-to-speech generation using parallel audio streams and an inner monologue of time-aligned text tokens.
citing papers explorer
-
ProtoSSL: Interpretable Prototype Learning from Unlabeled Time-Series Data
ProtoSSL discovers generalizable prototypes from unlabeled time-series via self-supervision and assigns them to new tasks for interpretable predictions, outperforming supervised baselines in low-data regimes on ECG datasets.
-
I'm Fine, But My Voice Isn't: Cross-Modal Affective Dissonance Detection for Reflective Journaling
A framework and model detect mismatches between text and voice emotions in journaling using a controlled TTS dataset and asymmetric attention architecture, achieving macro-F1 of 0.711.
-
Moshi: a speech-text foundation model for real-time dialogue
Moshi is the first real-time full-duplex spoken large language model that casts dialogue as speech-to-speech generation using parallel audio streams and an inner monologue of time-aligned text tokens.