Full-Duplex Dialogue, Smarter TTS, and Faster Voice Agents

As human--AI interactions become more conversational, full-duplex speech language models capable of natural real-time dialogue are growing in importance. Beyond generating appropriate responses, these models must coordinate turn-taking, backchanneling, and floor management in real time. Reinforcement learning (RL) provides a way to refine these behaviors through direct feedback on interaction outcomes. However, existing RL methods either apply timing feedback to a token policy or optimize semantic content, leaving the joint improvement of timing and content unresolved. We introduce HiPLEX, an RL framework that factorizes a pretrained full-duplex text policy into a control policy that decides when to emit content and a conditional content policy that decides what to emit. The first factor selects among 'pad', 'epad', and 'con'. The second selects a token only when 'con' is chosen. This hierarchy describes conditional actions within each frame and uses the model's existing text head. We route timing advantages to the token-group factor through event-causal masks derived from generated speech episodes, and route an LLM-judge semantic advantage to the conditional content factor. Across three Moshi seeds on Full-Duplex-Bench v1, HiPLEX reduces takeover rates during natural user pauses and backchannel opportunities, and shortens post-interruption response latency relative to GRPO, while maintaining comparable judged interruption-response quality. On Moshi and PersonaPlex, HiPLEX better matches pooled human turn-timing and backchannel-rate marginals than GRPO.

Full-duplex spoken dialogue models listen and speak at the same time, enabling voice agents to have natural, low-latency interactions that turn-based systems cannot offer. However, they are commonly evaluated against single-sided interlocutors: pre-recorded audio that cannot react, or an automated examiner that reacts in real time but only administers a fixed sequence of tests and is never graded. These single-sided frameworks evaluate only half of a two-body problem, where turn-taking, overlap, and interruption are joint products of two coupled speakers. We propose DyaFDB, a framework that evaluates full-duplex models in a dyadic setup: two models converse directly under assigned roles with cooperative or conflicting goals, and both sides are scored offline with an external judge. DyaFDB probes how the two models behave toward each other, such as how they take turns or carry an assigned role under different interests. We instantiate four tasks as 140 scenarios and record 7,560 conversations, covering six self- and cross-play pairings. Throughout the experiments, we observe that how a model behaves continually reshapes its partner. We thus demonstrate that each model must be both the examiner and examinee of the other, and no single fixed interlocutor can play both parts. We will release the scenarios, role prompts, and recording protocols between two full-duplex models, without any pre-recorded audio.

Large speech language models have demonstrated strong capabilities in unified cross-modal understanding and generation, yet paralinguistic cues, especially emotion, remain difficult to preserve. Existing systems typically rely on entangled acoustic representations, which allow the underlying language model to depend excessively on recovered lexical content instead of grounding its behavior in acoustic-prosodic evidence. We address this limitation with EMODE, an emotion-aware speech language model built around \textbf{Dynamic Para-Semantic Experts (DPSE)}. DPSE decomposes continuous speech features into semantic and paralinguistic pathways, routes them dynamically, and fuses them before integration into the language model. To turn this structural decomposition into functional specialization, EMODE is trained with a three-stage curriculum consisting of semantic warm-up, paralinguistic activation, and joint refinement, guided by Orthogonal Expert Guidance (OEG), Semantic-to-Acoustic Alignment (SAA), and Gating Diversity Regularization (GDR). Experiments on SER test, empathetic response evaluation, and the newly constructed bilingual MEPA benchmark show that EMODE improves the balance between lexical fidelity and emotional sensitivity, strengthens affect-grounded response generation, and exposes the value of explicit para-semantic factorization for robust cross-corpus emotion understanding.

Large audio language models answer questions about speech, sound, and music, yet their accuracy drops sharply on tasks that need fine-grained acoustic analysis. Judging which of two speakers has the higher pitch demands iterative signal-level reasoning that a content question does not. Current models spend the same computational depth on both. We introduce AdaLoop, a lightweight recurrent module that learns how many latent refinement steps a given audio--question pair requires. A shared transformer block iterates over the audio representation, guided by the question, while a learned halting mechanism exits the loop once the representation is ready. AdaLoop adds fewer than 3\% of the base model's parameters and plugs into any audio encoder--language model pair without modifying either component. Evaluated on three architecturally distinct models across MMSU, MMAU-Pro, and MMAR, AdaLoop raises the average accuracy by 2.9 to 3.8 points, with the largest gains on perception-heavy subtasks where the model learns to apply deeper reasoning.

Speech-LLMs often exhibit prompt overfitting, where models solely trained on automatic speech recognition (ASR) instruction fail to generalize to new instructions such as speech translation and continue to behave primarily as ASR system. We propose DirectSpeech2LLM, a simple end-to-end framework that preserves the instruction-following ability of the LLM on unseen tasks when conditioned on speech. It computes distance-based CTC loss over the frozen LLM embedding matrix and uses greedy CTC labels to derive geometrically and temporally aligned speech embeddings respectively as an input to the LLM. Trained solely on 960 hours of LibriSpeech ASR data, DirectSpeech2LLM outperforms the cascaded system on ASR (seen task) and generalizes zero-shot to speech translation and emotion recognition (two unseen tasks), closely matching the cascaded system upper bound on these two new instructions despite seeing neither during training. We also find that geometric alignment strength plays a smaller role than previously assumed, as our modified CTC loss is shown to provide sufficient implicit geometric grounding without requiring an explicit regression loss. Results are consistent across two LLM families and scale with both more training data and model capacity.

Accent leakage remains a critical challenge in cross-lingual zero-shot text-to-speech (TTS), where models inadvertently transfer both the speaker's timbre and source-language accent into the target-language output. In mask-reconstruction TTS this is amplified by a train--inference mismatch: training reconstructs from same-language context, while inference conditions on cross-lingual context. We close this gap by concatenating utterances from two languages and placing the reconstruction mask relative to the language boundary, so training presents the cross-lingual prompting pattern the model encounters at inference. This requires no language IDs, accent labels, parallel same-speaker recordings, or architectural changes---only the mask geometry differs. In a listening study with native speakers, region-aware masking reduces perceived accent from 4.3 to 0.5 on a 0-5 scale against the unadapted baseline, and by 18-51\% in seen regimes and 35-55\% on unseen source languages against bilingual adaptation alone, with no measurable loss in intelligibility or naturalness. Comparing mask geometries shows that placement, not the amount masked, governs the balance between accent suppression and speaker preservation: masks confined to the reconstruction region suppress accent most but retain the reference speaker least reliably, whereas two-region masking achieves both. Mask geometry is thus a simple, effective lever for accent robustness.

Speech becomes less intelligible in noisy environments, and humans naturally adapt their voice to compensate. Inspired by this behavior, we investigate whether a text-to-speech (TTS) model can be guided to produce more intelligible speech using activation steering, without retraining. We focus on two characteristics of the Lombard effect: increased vocal effort and hyper-articulation. We introduce a prompt-relative steering mechanism that prevents steering effects from accumulating during generation while allowing their strength to be adjusted dynamically. Across seen and unseen speakers and multiple languages, our method produces systematic changes in Lombard-related acoustic features, preserves speaker similarity (89-95%), and reduces WER under background noise by 7-22% at 1 dB SNR. These results show that pretrained TTS models can be dynamically controlled to generate more intelligible speech without retraining.

Tool-augmented speech assistants typically serialize automatic speech recognition, large language model inference, and external tool execution. As a result, tool latency is incurred only after the user has finished speaking and the LLM has identified the required tool calls. We present speculative tool execution for on-device cascaded voice agents, which predicts tool requests from partial ASR hypotheses and initiates tool execution while speech is still being received, thereby reducing end-to-end response latency. Our approach introduces a Predictor module that anticipates tool calls during speech recognition, executes them speculatively, and caches the results. The cached outputs are then injected into the LLM prompt, enabling faster responses. Additionally, to mitigate errors caused by user self-corrections during speech, we employ a rule-based validation mechanism that selectively injects only valid cached results. As a final safeguard, the LLM retains the ability to issue tool calls directly, ensuring that the latency of our framework is upper-bounded by the baseline serial execution pipeline in the worst case. We evaluate our method using live measurements from a fully implemented Android voice assistant. Our approach reduces the median time-to-first-audio from 5.79,s to 4.60,s and decreases the standard deviation from 3.49,s to 2.81,s, resulting in more predictable response latency.