Accurate segmentation of ultrasound images is essential for reliable medical diagnoses but is challenged by poor image quality and scarce labeled data. Prior approaches have relied on manually designed, complex network architectures to improve multi-scale feature extraction. However, such handcrafted models offer limited gains when prior knowledge is inadequate and are prone to overfitting on small datasets. In this paper, we introduce DeNAS-ViT, a Data efficient NAS-optimized Vision Transformer, the first method to leverage neural architecture search (NAS) for ultrasound image segmentation by automatically optimizing model architecture through token-level search. Specifically, we propose an efficient NAS module that performs multi-scale token search prior to the ViT’s attention mechanism, effectively capturing both contextual and local features while minimizing computational costs. Given ultrasound’s data scarcity and NAS’s inherent data demands, we further develop a NAS-guided semi-supervised learning (SSL) framework. This approach integrates network independence and contrastive learning within a stage-wise optimization strategy, significantly enhancing model robustness under limited-data conditions. Extensive experiments on public datasets demonstrate that DeNAS-ViT achieves state-of-the-art performance, maintaining robustness with minimal labeled data. Moreover, we highlight DeNAS-ViT’s generalization potential beyond ultrasound imaging, underscoring its broader applicability.

Inertial tracking is vital for robotic IoT and has gained popularity thanks to the ubiquity of low-cost inertial measurement units and deep learning-powered tracking algorithms. Existing works, however, have not fully utilized IMU measurements, particularly magnetometers, nor have they maximized the potential of deep learning to achieve the desired accuracy. To address these limitations, we introduceNeurIT, which elevates tracking accuracy to a new level.NeurITemploys a Time-Frequency Block-recurrent Transformer (TF-BRT) at its core, combining both RNN and Transformer to learn representative features in both time and frequency domains. To fully utilize IMU information, we strategically employ body-frame differentiation of magnetometers, considerably reducing the tracking error. We implementNeurITon a customized robotic platform and conduct evaluation in various indoor environments. Experimental results demonstrate thatNeurITachieves a mere 1-meter tracking error over a 300-meter distance. Notably, it significantly outperforms state-of-the-art baselines by 48.21% on unseen data. Moreover,NeurITdemonstrates robustness in large urban complexes and performs comparably to the visual-inertial approach (Tango Phone) in vision-favored conditions while surpassing it in feature-sparse settings. We believeNeurITtakes an important step forward toward practical neural inertial tracking for ubiquitous and scalable tracking of robotic things.NeurITis open-sourced here:https://github.com/aiot-lab/NeurIT.

Haoyang Su, Renqi Chen, Shixiang Tang, Zhenfei Yin, Xinzhe Zheng, Jinzhe Li, Biqing Qi, Qi Wu, Hui Li, Wanli Ouyang, Philip Torr, Bowen Zhou, Nanqing Dong. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2025.

Inertial tracking is vital for autonomous robots and has gained popularity with the ubiquity of low-cost Inertial Measurement Units (IMUs) and deep learning-powered tracking algorithms. Existing works, however, have not fully utilized IMU measurements, particularly magnetometers, nor maximized the potential of deep learning to achieve the desired accuracy. To bridge the gap, we introduce NeurIT, which employs a Time-Frequency Block-recurrent Transformer (TF-BRT) at its core, combining RNN and Transformer to learn both time-frequency representative features. To fully utilize IMU information, we strategically employ differentiation of body-frame magnetometers for orientation calibration in a sensor fusion manner. Experiments conducted in diverse environments show that NeurIT maintains a mere 1 -meter tracking error over a 300 - meter distance, surpassing state-of-the-art baselines by 48.21 % on unseen data. NeurIT also performs comparably to the visual-inertial approach (Tango Phone) in vision-favored conditions and surpasses it in plain environments. We share the code and data to promote further research: https://github.com/aiot-lab/NeurIT.

Deep learning-based computational methods have achieved promising results in predicting protein-protein interactions (PPIs). However, existing benchmarks predominantly focus on isolated pairwise evaluations, overlooking a model's capability to reconstruct biologically meaningful PPI networks, which is crucial for biology research. To address this gap, we introduce PRING, the first comprehensive benchmark that evaluates PRotein-protein INteraction prediction from a Graph-level perspective. PRING curates a high-quality, multi-species PPI network dataset comprising 21,484 proteins and 186,818 interactions, with well-designed strategies to address both data redundancy and leakage. Building on this golden-standard dataset, we establish two complementary evaluation paradigms: (1) topology-oriented tasks, which assess intra and cross-species PPI network construction, and (2) function-oriented tasks, including protein complex pathway prediction, GO module analysis, and essential protein justification. These evaluations not only reflect the model's capability to understand the network topology but also facilitate protein function annotation, biological module detection, and even disease mechanism analysis. Extensive experiments on four representative model categories, consisting of sequence similarity-based, naive sequence-based, protein language model-based, and structure-based approaches, demonstrate that current PPI models have potential limitations in recovering both structural and functional properties of PPI networks, highlighting the gap in supporting real-world biological applications. We believe PRING provides a reliable platform to guide the development of more effective PPI prediction models for the community. The dataset and source code of PRING are available at https://github.com/SophieSarceau/PRING.

The integration of Large Language Models (LLMs) with mobile devices is set to transform mental health care accessibility and quality. This paper introduces MindGuard, an autonomous LLM agent that utilizes mobile sensor data and engages in proactive, personalized conversations while ensuring user privacy through local processing. Unlike traditional mental health AI tools, MindGuard enables real-time, context-aware interventions by dynamically adapting to users' emotional and physiological states. The real-world implementation demonstrates its effectiveness with the ultimate goal of creating an accessible, scalable, and personalized mental healthcare ecosystem for anyone with smart mobile devices.

Predicting blood glucose levels is fundamental for precise primary care of type-1 diabetes (T1D) patients. However, it is challenging to predict glucose levels accurately, not to mention the early alarm of adverse events (hyperglycemia and hypoglycemia), namely the minority class. In this paper, we propose BG-BERT, a novel self-supervised learning framework for blood glucose level prediction. In particular, BG-BERT incorporates masked autoencoder to capture rich contextual information of blood glucose records for accurate prediction. More specifically, SMOTE data augmentation and shrinkage loss are employed to effectively handle adverse events without discrimination. We evaluate BG-BERT on two benchmark datasets against two state-of-the-art base-line models. The experimental results highlight the significant improvements achieved by BG-BERT in glucose level prediction accuracy (measured by RMSE) and sensitivity to adverse events, with average lifting ratios of 9.5% and 44.9%, respectively.

Large language models (LLMs) have demonstrated remarkable efficacy across knowledge-intensive tasks. Nevertheless, their untapped potential in crop science presents an opportunity for advancement. To narrow this gap, we introduce CROP, which includes a novel instruction tuning dataset specifically designed to enhance LLMs’ professional capabilities in the crop science sector, along with a benchmark that serves as a comprehensive evaluation of LLMs’ understanding of the domain knowledge. The CROP dataset is curated through a task-oriented and LLM-human integrated pipeline, comprising 210,038 single-turn and 1,871 multi-turn dialogues related to crop science scenarios. The CROP benchmark includes 5,045 multiple-choice questions covering three difficulty levels. Our experiments based on the CROP benchmark demonstrate notable enhancements in crop science-related tasks when LLMs are fine-tuned with the CROP dataset. To the best of our knowledge, CROP dataset is the first-ever instruction tuning dataset in the crop science domain. We anticipate that CROP will accelerate the adoption of LLMs in the domain of crop science, ultimately contributing to global food production.