arXiv subjects

Tuochao Chen

Publications and source records attributed to Tuochao Chen.

16 recordsLinked to original sources

arXiv2026

Micro Language Models Enable Instant Responses

Edge devices such as smartwatches and smart glasses cannot continuously run even the smallest 100M-1B parameter language models due to power and compute constraints, yet cloud inference introduces multi-second latencies that break the illusion of a responsive assistant. We introduce micro language models ($\mu$LMs): ultra-compact models (8M-30M parameters) that instantly generate the first 4-8 words of a contextually grounded response on-device, while a cloud model completes it; thus, masking the cloud latency. We show that useful language generation survives at this extreme scale with our models matching several 70M-256M-class existing models. We design a collaborative generation framework that reframes the cloud model as a continuator rather than a respondent, achieving seamless mid-sentence handoffs and structured graceful recovery via three error correction methods when the local opener goes wrong. Empirical results show that $\mu$LMs can initiate responses that larger models complete seamlessly, demonstrating that orders-of-magnitude asymmetric collaboration is achievable and unlocking responsive AI for extremely resource-constrained devices. The model checkpoint and demo are available at https://github.com/Sensente/micro_language_model_swen_project.

cs.CL↗

arXiv2025

AV-Dialog: Spoken Dialogue Models with Audio-Visual Input

Dialogue models falter in noisy, multi-speaker environments, often producing irrelevant responses and awkward turn-taking. We present AV-Dialog, the first multimodal dialog framework that uses both audio and visual cues to track the target speaker, predict turn-taking, and generate coherent responses. By combining acoustic tokenization with multi-task, multi-stage training on monadic, synthetic, and real audio-visual dialogue datasets, AV-Dialog achieves robust streaming transcription, semantically grounded turn-boundary detection and accurate responses, resulting in a natural conversational flow. Experiments show that AV-Dialog outperforms audio-only models under interference, reducing transcription errors, improving turn-taking prediction, and enhancing human-rated dialogue quality. These results highlight the power of seeing as well as hearing for speaker-aware interaction, paving the way for {spoken} dialogue agents that perform {robustly} in real-world, noisy environments.

cs.CL↗

arXiv2025

Proactive Hearing Assistants that Isolate Egocentric Conversations

We introduce proactive hearing assistants that automatically identify and separate the wearer's conversation partners, without requiring explicit prompts. Our system operates on egocentric binaural audio and uses the wearer's self-speech as an anchor, leveraging turn-taking behavior and dialogue dynamics to infer conversational partners and suppress others. To enable real-time, on-device operation, we propose a dual-model architecture: a lightweight streaming model runs every 12.5 ms for low-latency extraction of the conversation partners, while a slower model runs less frequently to capture longer-range conversational dynamics. Results on real-world 2- and 3-speaker conversation test sets, collected with binaural egocentric hardware from 11 participants totaling 6.8 hours, show generalization in identifying and isolating conversational partners in multi-conversation settings. Our work marks a step toward hearing assistants that adapt proactively to conversational dynamics and engagement. More information can be found on our website: https://proactivehearing.cs.washington.edu/

cs.CL↗

arXiv2025

Reading Between the Lines: The One-Sided Conversation Problem

Conversational AI is constrained in many real-world settings where only one side of a dialogue can be recorded, such as telemedicine, call centers, and smart glasses. We formalize this as the one-sided conversation problem (1SC): inferring and learning from one side of a conversation. We study two tasks: (1) reconstructing the missing speaker's turns for real-time use cases, and (2) generating summaries from one-sided transcripts. Evaluating prompting and finetuned models on MultiWOZ, DailyDialog, and Candor with both human A/B testing and LLM-as-a-judge metrics, we find that access to one future turn and information about utterance length improves reconstruction, placeholder prompting helps to mitigate hallucination, and while large models generate promising reconstructions with prompting, smaller models require finetuning. Further, high-quality summaries can be generated without reconstructing missing turns. We present 1SC as a novel challenge and report promising results that mark a step toward privacy-aware conversational AI.

cs.CL↗

arXiv2025

Wireless Hearables With Programmable Speech AI Accelerators

The conventional wisdom has been that designing ultra-compact, battery-constrained wireless hearables with on-device speech AI models is challenging due to the high computational demands of streaming deep learning models. Speech AI models require continuous, real-time audio processing, imposing strict computational and I/O constraints. We present NeuralAids, a fully on-device speech AI system for wireless hearables, enabling real-time speech enhancement and denoising on compact, battery-constrained devices. Our system bridges the gap between state-of-the-art deep learning for speech enhancement and low-power AI hardware by making three key technical contributions: 1) a wireless hearable platform integrating a speech AI accelerator for efficient on-device streaming inference, 2) an optimized dual-path neural network designed for low-latency, high-quality speech enhancement, and 3) a hardware-software co-design that uses mixed-precision quantization and quantization-aware training to achieve real-time performance under strict power constraints. Our system processes 6 ms audio chunks in real-time, achieving an inference time of 5.54 ms while consuming 71.6 mW. In real-world evaluations, including a user study with 28 participants, our system outperforms prior on-device models in speech quality and noise suppression, paving the way for next-generation intelligent wireless hearables that can enhance hearing entirely on-device.

cs.SD↗

arXiv2025

TF-MLPNet: Tiny Real-Time Neural Speech Separation

Speech separation on hearable devices can enable transformative augmented and enhanced hearing capabilities. However, state-of-the-art speech separation networks cannot run in real-time on tiny, low-power neural accelerators designed for hearables, due to their limited compute capabilities. We present TF-MLPNet, the first speech separation network capable of running in real-time on such low-power accelerators while outperforming existing streaming models for blind speech separation and target speech extraction. Our network operates in the time-frequency domain, processing frequency sequences with stacks of fully connected layers that alternate along the channel and frequency dimensions, and independently processing the time sequence at each frequency bin using convolutional layers. Results show that our mixed-precision quantization-aware trained (QAT) model can process 6 ms audio chunks in real-time on the GAP9 processor, achieving a 3.5-4x runtime reduction compared to prior speech separation models.

cs.SD↗

arXiv2025

LLAMAPIE: Proactive In-Ear Conversation Assistants

We introduce LlamaPIE, the first real-time proactive assistant designed to enhance human conversations through discreet, concise guidance delivered via hearable devices. Unlike traditional language models that require explicit user invocation, this assistant operates in the background, anticipating user needs without interrupting conversations. We address several challenges, including determining when to respond, crafting concise responses that enhance conversations, leveraging knowledge of the user for context-aware assistance, and real-time, on-device processing. To achieve this, we construct a semi-synthetic dialogue dataset and propose a two-model pipeline: a small model that decides when to respond and a larger model that generates the response. We evaluate our approach on real-world datasets, demonstrating its effectiveness in providing helpful, unobtrusive assistance. User studies with our assistant, implemented on Apple Silicon M2 hardware, show a strong preference for the proactive assistant over both a baseline with no assistance and a reactive model, highlighting the potential of LlamaPie to enhance live conversations.

cs.LG↗

arXiv2025

SoundSculpt: Direction and Semantics Driven Ambisonic Target Sound Extraction

This paper introduces SoundSculpt, a neural network designed to extract target sound fields from ambisonic recordings. SoundSculpt employs an ambisonic-in-ambisonic-out architecture and is conditioned on both spatial information (e.g., target direction obtained by pointing at an immersive video) and semantic embeddings (e.g., derived from image segmentation and captioning). Trained and evaluated on synthetic and real ambisonic mixtures, SoundSculpt demonstrates superior performance compared to various signal processing baselines. Our results further reveal that while spatial conditioning alone can be effective, the combination of spatial and semantic information is beneficial in scenarios where there are secondary sound sources spatially close to the target. Additionally, we compare two different semantic embeddings derived from a text description of the target sound using text encoders.

eess.AS↗

arXiv2025

Spatial Speech Translation: Translating Across Space With Binaural Hearables

Imagine being in a crowded space where people speak a different language and having hearables that transform the auditory space into your native language, while preserving the spatial cues for all speakers. We introduce spatial speech translation, a novel concept for hearables that translate speakers in the wearer's environment, while maintaining the direction and unique voice characteristics of each speaker in the binaural output. To achieve this, we tackle several technical challenges spanning blind source separation, localization, real-time expressive translation, and binaural rendering to preserve the speaker directions in the translated audio, while achieving real-time inference on the Apple M2 silicon. Our proof-of-concept evaluation with a prototype binaural headset shows that, unlike existing models, which fail in the presence of interference, we achieve a BLEU score of up to 22.01 when translating between languages, despite strong interference from other speakers in the environment. User studies further confirm the system's effectiveness in spatially rendering the translated speech in previously unseen real-world reverberant environments. Taking a step back, this work marks the first step towards integrating spatial perception into speech translation.

cs.CL↗

arXiv2024

Target conversation extraction: Source separation using turn-taking dynamics

Extracting the speech of participants in a conversation amidst interfering speakers and noise presents a challenging problem. In this paper, we introduce the novel task of target conversation extraction, where the goal is to extract the audio of a target conversation based on the speaker embedding of one of its participants. To accomplish this, we propose leveraging temporal patterns inherent in human conversations, particularly turn-taking dynamics, which uniquely characterize speakers engaged in conversation and distinguish them from interfering speakers and noise. Using neural networks, we show the feasibility of our approach on English and Mandarin conversation datasets. In the presence of interfering speakers, our results show an 8.19 dB improvement in signal-to-noise ratio for 2-speaker conversations and a 7.92 dB improvement for 2-4-speaker conversations. Code, dataset available at https://github.com/chentuochao/Target-Conversation-Extraction.

cs.CL↗

arXiv2024

Knowledge boosting during low-latency inference

Models for low-latency, streaming applications could benefit from the knowledge capacity of larger models, but edge devices cannot run these models due to resource constraints. A possible solution is to transfer hints during inference from a large model running remotely to a small model running on-device. However, this incurs a communication delay that breaks real-time requirements and does not guarantee that both models will operate on the same data at the same time. We propose knowledge boosting, a novel technique that allows a large model to operate on time-delayed input during inference, while still boosting small model performance. Using a streaming neural network that processes 8 ms chunks, we evaluate different speech separation and enhancement tasks with communication delays of up to six chunks or 48 ms. Our results show larger gains where the performance gap between the small and large models is wide, demonstrating a promising method for large-small model collaboration for low-latency applications. Code, dataset, and audio samples available at https://knowledgeboosting.cs.washington.edu/.

cs.LG↗

arXiv2024

Look Once to Hear: Target Speech Hearing with Noisy Examples

In crowded settings, the human brain can focus on speech from a target speaker, given prior knowledge of how they sound. We introduce a novel intelligent hearable system that achieves this capability, enabling target speech hearing to ignore all interfering speech and noise, but the target speaker. A naive approach is to require a clean speech example to enroll the target speaker. This is however not well aligned with the hearable application domain since obtaining a clean example is challenging in real world scenarios, creating a unique user interface problem. We present the first enrollment interface where the wearer looks at the target speaker for a few seconds to capture a single, short, highly noisy, binaural example of the target speaker. This noisy example is used for enrollment and subsequent speech extraction in the presence of interfering speakers and noise. Our system achieves a signal quality improvement of 7.01 dB using less than 5 seconds of noisy enrollment audio and can process 8 ms of audio chunks in 6.24 ms on an embedded CPU. Our user studies demonstrate generalization to real-world static and mobile speakers in previously unseen indoor and outdoor multipath environments. Finally, our enrollment interface for noisy examples does not cause performance degradation compared to clean examples, while being convenient and user-friendly. Taking a step back, this paper takes an important step towards enhancing the human auditory perception with artificial intelligence. We provide code and data at: https://github.com/vb000/LookOnceToHear.

cs.SD↗

arXiv2023

Underwater 3D positioning on smart devices

The emergence of water-proof mobile and wearable devices (e.g., Garmin Descent and Apple Watch Ultra) designed for underwater activities like professional scuba diving, opens up opportunities for underwater networking and localization capabilities on these devices. Here, we present the first underwater acoustic positioning system for smart devices. Unlike conventional systems that use floating buoys as anchors at known locations, we design a system where a dive leader can compute the relative positions of all other divers, without any external infrastructure. Our intuition is that in a well-connected network of devices, if we compute the pairwise distances, we can determine the shape of the network topology. By incorporating orientation information about a single diver who is in the visual range of the leader device, we can then estimate the positions of all the remaining divers, even if they are not within sight. We address various practical problems including detecting erroneous distance estimates, addressing rotational and flipping ambiguities as well as designing a distributed timestamp protocol that scales linearly with the number of devices. Our evaluations show that our distributed system running on underwater deployments of 4-5 commodity smart devices can perform pairwise ranging and localization with median errors of 0.5-0.9 m and 0.9-1.6 m

cs.NI↗

arXiv2023

Real-Time Target Sound Extraction

We present the first neural network model to achieve real-time and streaming target sound extraction. To accomplish this, we propose Waveformer, an encoder-decoder architecture with a stack of dilated causal convolution layers as the encoder, and a transformer decoder layer as the decoder. This hybrid architecture uses dilated causal convolutions for processing large receptive fields in a computationally efficient manner while also leveraging the generalization performance of transformer-based architectures. Our evaluations show as much as 2.2-3.3 dB improvement in SI-SNRi compared to the prior models for this task while having a 1.2-4x smaller model size and a 1.5-2x lower runtime. We provide code, dataset, and audio samples: https://waveformer.cs.washington.edu/.

cs.SD↗

arXiv2022

Underwater Acoustic Ranging Between Smartphones

We present a novel underwater system that can perform acoustic ranging between commodity smartphones. To achieve this, we design a real-time underwater ranging protocol that computes the time-of-flight between smartphones. To address the severe underwater multipath, we present a dual-microphone optimization algorithm that can more reliably identify the direct path. Our underwater evaluations show that our system has median errors of 0.48-0.86 m at distances upto 35 m. Further, our system can operate across smartphone model pairs and works in the presence of clock drifts. While existing underwater localization research is targeted for custom hydrophone hardware, we believe that our work breaks new ground by demonstrating a path to bringing underwater ranging capabilities to billions of existing smartphones, without additional hardware.

cs.NI↗

arXiv2022

Underwater Messaging Using Mobile Devices

Since its inception, underwater digital acoustic communication has required custom hardware that neither has the economies of scale nor is pervasive. We present the first acoustic system that brings underwater messaging capabilities to existing mobile devices like smartphones and smart watches. Our software-only solution leverages audio sensors, i.e., microphones and speakers, ubiquitous in today's devices to enable acoustic underwater communication between mobile devices. To achieve this, we design a communication system that in real-time adapts to differences in frequency responses across mobile devices, changes in multipath and noise levels at different locations and dynamic channel changes due to mobility. We evaluate our system in six different real-world underwater environments with depths of 2-15 m in the presence of boats, ships and people fishing and kayaking. Our results show that our system can in real-time adapt its frequency band and achieve bit rates of 100 bps to 1.8 kbps and a range of 30 m. By using a lower bit rate of 10-20 bps, we can further increase the range to 100 m. As smartphones and watches are increasingly being used in underwater scenarios, our software-based approach has the potential to make underwater messaging capabilities widely available to anyone with a mobile device. Project page with open-source code and data can be found here: https://underwatermessaging.cs.washington.edu/

cs.NI↗