arXiv subjects

Artem Dementyev

Publications and source records attributed to Artem Dementyev.

3 recordsLinked to original sources

arXiv2026

PhaseCoder: Microphone Geometry-Agnostic Spatial Audio Understanding for Multimodal LLMs

Current multimodal LLMs process audio as a mono stream, ignoring the rich spatial information essential for embodied AI. Existing spatial audio models, conversely, are constrained to fixed microphone geometries, preventing deployment across diverse devices. We present PhaseCoder, a transformer-only spatial audio encoder that is agnostic to microphone geometry. PhaseCoder takes raw multichannel audio and microphone coordinates as inputs to perform localization and produces robust spatial embeddings. We demonstrate that Gemma 3n LLM can be fine-tuned to reason over "Spatial Audio Tokens" produced by PhaseCoder. We show our encoder achieves state-of-the-art results on microphone-invariant localization benchmarks and, for the first time, enables an LLM to perform complex spatial reasoning and targeted transcription tasks from an arbitrary microphone array.

cs.SD↗

arXiv2025

Towards Sub-millisecond Latency Real-Time Speech Enhancement Models on Hearables

Low latency models are critical for real-time speech enhancement applications, such as hearing aids and hearables. However, the sub-millisecond latency space for resource-constrained hearables remains underexplored. We demonstrate speech enhancement using a computationally efficient minimum-phase FIR filter, enabling sample-by-sample processing to achieve mean algorithmic latency of 0.32 ms to 1.25 ms. With a single microphone, we observe a mean SI-SDRi of 4.1 dB. The approach shows generalization with a DNSMOS increase of 0.2 on unseen audio recordings. We use a lightweight LSTM-based model of 626k parameters to generate FIR taps. Using a real hardware implementation on a low-power DSP, our system can run with 376 MIPS and a mean end-to-end latency of 3.35 ms. In addition, we provide a comparison with existing low-latency spectral masking techniques. We hope this work will enable a better understanding of latency and can be used to improve the comfort and usability of hearables.

cs.SD↗

arXiv2025

SpeechCompass: Enhancing Mobile Captioning with Diarization and Directional Guidance via Multi-Microphone Localization

Speech-to-text capabilities on mobile devices have proven helpful for hearing and speech accessibility, language translation, note-taking, and meeting transcripts. However, our foundational large-scale survey (n=263) shows that the inability to distinguish and indicate speaker direction makes them challenging in group conversations. SpeechCompass addresses this limitation through real-time, multi-microphone speech localization, where the direction of speech allows visual separation and guidance (e.g., arrows) in the user interface. We introduce efficient real-time audio localization algorithms and custom sound perception hardware running on a low-power microcontroller and four integrated microphones, which we characterize in technical evaluations. Informed by a large-scale survey (n=494), we conducted an in-person study of group conversations with eight frequent users of mobile speech-to-text, who provided feedback on five visualization styles. The value of diarization and visualizing localization was consistent across participants, with everyone agreeing on the value and potential of directional guidance for group conversations.

cs.HC↗