Lujia Yang, Weicai Yan, Yongbo He, Qifei Zhang, Tao Jin, Jinshan Zhang, Meng Xi, Jianwei Yin. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2026.

Extensive research on LLM-based spoken dialogue systems has significantly advanced the development of intelligent voice assistants.However, the integration of role information within speech remains an underexplored area, limiting its application in real-world scenarios, particularly in multi-party dialogue settings.With the growing demand for personalization, voice assistants that can recognize and remember users establish a deeper connection with them.We focus on enabling LLMs with speaker-awareness capabilities and enhancing their understanding of character settings through synthetic data to generate contextually appropriate responses.We introduce Persona-Dialogue, the first large-scale multi-party spoken dialogue dataset that incorporates speaker profiles.Based on this dataset, we propose PAChat, an architecture that simultaneously models both linguistic content and speaker features, allowing LLMs to map character settings to speaker identities in speech.Through extensive experiments, we demonstrate that PAChat successfully achieves speaker-specific responses, character understanding, and the generation of targeted replies in multi-party dialogue scenarios, surpassing existing spoken dialogue systems.For more details, please visit our demo page at https

Estimating camera motion and continuously reconstructing dense scenes in deformable environments presents a complex and open challenge. Many existing approaches tend to rely on assumptions about the scene’s topology or the nature of deformable motion. However, these assumptions do not hold true in medical endoscopy applications. To address these challenges, we introduce DDS-SLAM, a novel dense deformable semantic neural SLAM that achieves accurate camera tracking, continuous dense scene reconstruction, and high-quality image rendering in deformable scenes. First, we propose a novel hybrid neural scene representation method capable of capturing both natural and artificial deformations. Additionally, by leveraging the 2D semantic information of the scene, we introduce a semantic loss function based on semantic distance fields. This approach guides network optimization at a higher level, thereby enhancing system performance. Furthermore, we validate our method through a series of experiments conducted on several representative medical datasets, demonstrating its superiority over other state-of-the-art approaches. The code is available at: https://github.com/IRMVLab/DDS-SLAM.

This study aimed to identify and assess vegetation indices (VIs) and their optimal band combinations using unmanned aerial vehicle (UAV) hyperspectral imagery for the quantitative inversion of wheat stripe rust. This would offer guidance for selecting rust-resistant phenotypes and facilitate large-scale disease monitoring using aerial and spaceborne remote sensing images. The experimental design encompassed 960 wheat varieties (strains) in agricultural fields. Hyperspectral imagery was acquired at 100m altitude during different disease stages, and disease index (DI) was investigated per plot. A custom program explored VIs with two, three, and four bands using 30 calculation methods and 3,463,790 band combinations. Regression models employed three-fold cross-validation and multilayer perceptron (MLP) algorithms, with the mean R2value indicating optimal index and band combinations. The results revealed that the chosen VIs were effective in inverting the DI. Selected two-band VIs included MGRVI (531, 571), with R2=0.746±0.01618; the optimal three-band VIs was ARI2 (531, 550±10, 640±25), with R2=0.755±0.00896; and the best four-band VIs was DBSI (531, 551, 750, 799), with R2=0.778±0.01300, which was comparable to full-band modeling (R2=0.775±0.01508). The models’ performance improved with an increasing number of bands in the VIs. This study demonstrated that appropriate multi-VIs modeling enhances performance compared to single-VIs modeling, e.g., six combinations of VIs achieved R2=0.790±0.01141. These findings underscore the potential of integrating machine learning algorithms and vegetation indices for quantifying wheat rust diseases, laying the foundation for developing airborne and spaceborne imaging sensors for large-scale wheat rust monitoring.