Enhancing Physical Consistency in Lightweight World Models

A major challenge in deploying world models is the trade-off between size and performance. Large world models can capture rich physical dynamics but require massive computing resources, making them impractical for edge devices. Small world models are easier to deploy but often struggle to learn accurate physics, leading to poor predictions. We propose the Physics-Informed BEV World Model (PIWM), a compact model designed to efficiently capture physical interactions in bird's-eye-view (BEV) representations. PIWM uses Soft Mask during training to improve dynamic object modeling and future prediction. We also introduce a simple yet effective technique, Warm Start, for inference to enhance prediction quality with a zero-shot model. Experiments show that at the same parameter scale (400M), PIWM surpasses the baseline by 60.6% in weighted overall score. Moreover, even when compared with the largest baseline model (400M), the smallest PIWM (130M Soft Mask) achieves a 7.4% higher weighted overall score with a 28% faster inference speed.

EfficientIML: Efficient High-Resolution Image Manipulation Localization

With imaging devices delivering ever-higher resolutions and the emerging diffusion-based forgery methods, current detectors trained only on traditional datasets (with splicing, copy-moving and object removal forgeries) lack exposure to this new manipulation type. To address this, we propose a novel high-resolution SIF dataset of 1200+ diffusion-generated manipulations with semantically extracted masks. However, this also imposes a challenge on existing methods, as they face significant computational resource constraints due to their prohibitive computational complexities. Therefore, we propose a novel EfficientIML model with a lightweight, three-stage EfficientRWKV backbone. EfficientRWKV's hybrid state-space and attention network captures global context and local details in parallel, while a multi-scale supervision strategy enforces consistency across hierarchical predictions. Extensive evaluations on our dataset and standard benchmarks demonstrate that our approach outperforms ViT-based and other SOTA lightweight baselines in localization performance, FLOPs and inference speed, underscoring its suitability for real-time forensic applications.

Bistatic Target Detection by Exploiting Both Deterministic Pilots and Unknown Random Data Payloads

Integrated sensing and communication (ISAC) plays a crucial role in 6G, to enable innovative applications such as drone surveillance, urban air mobility, and low-altitude logistics. However, the hybrid ISAC signal, which comprises deterministic pilot and random data payload components, poses challenges for target detection due to two reasons: 1) these two components cause coupled shifts in both the mean and variance of the received signal, and 2) the random data payloads are typically unknown to the sensing receiver in the bistatic setting. Unfortunately, these challenges could not be tackled by existing target detection algorithms. In this paper, a generalized likelihood ratio test (GLRT)-based detector is derived, by leveraging the known deterministic pilots and the statistical characteristics of the unknown random data payloads. Due to the analytical intractability of exact performance characterization, we perform an asymptotic analysis for the false alarm probability and detection probability of the proposed detector. The results highlight a critical trade-off: both deterministic and random components improve detection reliability, but the latter also brings statistical uncertainty that hinders detection performance. Simulations validate the theoretical findings and demonstrate the effectiveness of the proposed detector, which highlights the necessity of designing a dedicated detector to fully exploited the signaling resources assigned to random data payloads.

A Rapid Iterative Trajectory Planning Method for Automated Parking through Differential Flatness

As autonomous driving continues to advance, automated parking is becoming increasingly essential. However, significant challenges arise when implementing path velocity decomposition (PVD) trajectory planning for automated parking. The primary challenge is ensuring rapid and precise collision-free trajectory planning, which is often in conflict. The secondary challenge involves maintaining sufficient control feasibility of the planned trajectory, particularly at gear shifting points (GSP). This paper proposes a PVD-based rapid iterative trajectory planning (RITP) method to solve the above challenges. The proposed method effectively balances the necessity for time efficiency and precise collision avoidance through a novel collision avoidance framework. Moreover, it enhances the overall control feasibility of the planned trajectory by incorporating the vehicle kinematics model and including terminal smoothing constraints (TSC) at GSP during path planning. Specifically, the proposed method leverages differential flatness to ensure the planned path adheres to the vehicle kinematic model. Additionally, it utilizes TSC to maintain curvature continuity at GSP, thereby enhancing the control feasibility of the overall trajectory. The simulation results demonstrate superior time efficiency and tracking errors compared to model-integrated and other iteration-based trajectory planning methods. In the real-world experiment, the proposed method was implemented and validated on a ROS-based vehicle, demonstrating the applicability of the RITP method for real vehicles.

XEmoRAG: Cross-Lingual Emotion Transfer with Controllable Intensity Using Retrieval-Augmented Generation

Zero-shot emotion transfer in cross-lingual speech synthesis refers to generating speech in a target language, where the emotion is expressed based on reference speech from a different source language. However, this task remains challenging due to the scarcity of parallel multilingual emotional corpora, the presence of foreign accent artifacts, and the difficulty of separating emotion from language-specific prosodic features. In this paper, we propose XEmoRAG, a novel framework to enable zero-shot emotion transfer from Chinese to Thai using a large language model (LLM)-based model, without relying on parallel emotional data. XEmoRAG extracts language-agnostic emotional embeddings from Chinese speech and retrieves emotionally matched Thai utterances from a curated emotional database, enabling controllable emotion transfer without explicit emotion labels. Additionally, a flow-matching alignment module minimizes pitch and duration mismatches, ensuring natural prosody. It also blends Chinese timbre into the Thai synthesis, enhancing rhythmic accuracy and emotional expression, while preserving speaker characteristics and emotional consistency. Experimental results show that XEmoRAG synthesizes expressive and natural Thai speech using only Chinese reference audio, without requiring explicit emotion labels. These results highlight XEmoRAG's capability to achieve flexible and low-resource emotional transfer across languages. Our demo is available at https://tlzuo-lesley.github.io/Demo-page/ .

MSU-Bench: Towards Understanding the Conversational Multi-talker Scenarios

Spoken Language Understanding (SLU) has progressed from traditional single-task methods to large audio language model (LALM) solutions. Yet, most existing speech benchmarks focus on single-speaker or isolated tasks, overlooking the challenges posed by multi-speaker conversations that are common in real-world scenarios. We introduce MSU-Bench, a comprehensive benchmark for evaluating multi-speaker conversational understanding with a speaker-centric design. Our hierarchical framework covers four progressive tiers: single-speaker static attribute understanding, single-speaker dynamic attribute understanding, multi-speaker background understanding, and multi-speaker interaction understanding. This structure ensures all tasks are grounded in speaker-centric contexts, from basic perception to complex reasoning across multiple speakers. By evaluating state-of-the-art models on MSU-Bench, we demonstrate that as task complexity increases across the benchmark's tiers, all models exhibit a significant performance decline. We also observe a persistent capability gap between open-source models and closed-source commercial ones, particularly in multi-speaker interaction reasoning. These findings validate the effectiveness of MSU-Bench for assessing and advancing conversational understanding in realistic multi-speaker environments. Demos can be found in the supplementary material.

UniFlow: Unifying Speech Front-End Tasks via Continuous Generative Modeling

Generative modeling has recently achieved remarkable success across image, video, and audio domains, demonstrating powerful capabilities for unified representation learning. Yet speech front-end tasks such as speech enhancement (SE), target speaker extraction (TSE), acoustic echo cancellation (AEC), and language-queried source separation (LASS) remain largely tackled by disparate, task-specific solutions. This fragmentation leads to redundant engineering effort, inconsistent performance, and limited extensibility. To address this gap, we introduce UniFlow, a unified framework that employs continuous generative modeling to tackle diverse speech front-end tasks in a shared latent space. Specifically, UniFlow utilizes a waveform variational autoencoder (VAE) to learn a compact latent representation of raw audio, coupled with a Diffusion Transformer (DiT) that predicts latent updates. To differentiate the speech processing task during the training, learnable condition embeddings indexed by a task ID are employed to enable maximal parameter sharing while preserving task-specific adaptability. To balance model performance and computational efficiency, we investigate and compare three generative objectives: denoising diffusion, flow matching, and mean flow within the latent domain. We validate UniFlow on multiple public benchmarks, demonstrating consistent gains over state-of-the-art baselines. UniFlow's unified latent formulation and conditional design make it readily extensible to new tasks, providing an integrated foundation for building and scaling generative speech processing pipelines. To foster future research, we will open-source our codebase.

Llasa+: Free Lunch for Accelerated and Streaming Llama-Based Speech Synthesis

Recent progress in text-to-speech (TTS) has achieved impressive naturalness and flexibility, especially with the development of large language model (LLM)-based approaches. However, existing autoregressive (AR) structures and large-scale models, such as Llasa, still face significant challenges in inference latency and streaming synthesis. To deal with the limitations, we introduce Llasa+, an accelerated and streaming TTS model built on Llasa. Specifically, to accelerate the generation process, we introduce two plug-and-play Multi-Token Prediction (MTP) modules following the frozen backbone. These modules allow the model to predict multiple tokens in one AR step. Additionally, to mitigate potential error propagation caused by inaccurate MTP, we design a novel verification algorithm that leverages the frozen backbone to validate the generated tokens, thus allowing Llasa+ to achieve speedup without sacrificing generation quality. Furthermore, we design a causal decoder that enables streaming speech reconstruction from tokens. Extensive experiments show that Llasa+ achieves a 1.48X speedup without sacrificing generation quality, despite being trained only on LibriTTS. Moreover, the MTP-and-verification framework can be applied to accelerate any LLM-based model. All codes and models are publicly available at https://github.com/ASLP-lab/LLaSA_Plus.

EchoFree: Towards Ultra Lightweight and Efficient Neural Acoustic Echo Cancellation

In recent years, neural networks (NNs) have been widely applied in acoustic echo cancellation (AEC). However, existing approaches struggle to meet real-world low-latency and computational requirements while maintaining performance. To address this challenge, we propose EchoFree, an ultra lightweight neural AEC framework that combines linear filtering with a neural post filter. Specifically, we design a neural post-filter operating on Bark-scale spectral features. Furthermore, we introduce a two-stage optimization strategy utilizing self-supervised learning (SSL) models to improve model performance. We evaluate our method on the blind test set of the ICASSP 2023 AEC Challenge. The results demonstrate that our model, with only 278K parameters and 30 MMACs computational complexity, outperforms existing low-complexity AEC models and achieves performance comparable to that of state-of-the-art lightweight model DeepVQE-S. The audio examples are available.

REF-VC: Robust, Expressive and Fast Zero-Shot Voice Conversion with Diffusion Transformers

In real-world voice conversion applications, environmental noise in source speech and user demands for expressive output pose critical challenges. Traditional ASR-based methods ensure noise robustness but suppress prosody, while SSL-based models improve expressiveness but suffer from timbre leakage and noise sensitivity. This paper proposes REF-VC, a noise-robust expressive voice conversion system. Key innovations include: (1) A random erasing strategy to mitigate the information redundancy inherent in SSL feature, enhancing noise robustness and expressiveness; (2) Implicit alignment inspired by E2TTS to suppress non-essential feature reconstruction; (3) Integration of Shortcut Models to accelerate flow matching inference, significantly reducing to 4 steps. Experimental results demonstrate that our model outperforms baselines such as Seed-VC in zero-shot scenarios on the noisy set, while also performing comparably to Seed-VC on the clean set. In addition, REF-VC can be compatible with singing voice conversion within one model.