Beacon Tech (Israel)

In the context of stringent regulatory oversight and highly volatile demand, medical device supply chains (MDSCs) have long been plagued by high costs, high inventory levels, and slow response times. This paper integrates the closed-loop operation concept with a digital technology stack to construct a “three-end five-flow” closed-loop smart supply chain (CPSC) architecture. Based on 128 quarterly observations from 2021 to 2024 and over 100,000 UDI-level micro-data points, a mixed-method approach is employed to examine the efficiency improvement effects.

Because of predicting all the target tokens in parallel, the non-autoregressive models greatly improve the decoding efficiency of speech recognition compared with traditional autoregressive models. In this work, we present dynamic alignment Mask CTC, introducing two methods: (1) Aligned Cross Entropy (AXE), finding the monotonic alignment that minimizes the cross-entropy loss through dynamic programming, (2) Dynamic Rectification, creating new training samples by replacing some masks with model predicted tokens. The AXE ignores the absolute position alignment between prediction and ground truth sentence and focuses on tokens matching in relative order. The dynamic rectification method makes the model capable of simulating the non-mask but possible wrong tokens, even if they have high confidence. Our experiments on WSJ dataset demonstrated that not only AXE loss but also the rectification method could improve the WER performance of Mask CTC.

Non-parallel many-to-many voice conversion is a kind of style transfer task in speech. Recently, AutoVC has been applied in this field as a popular solution, as it can achieve distribution-matching style transfer by training only the re- construction loss. However, in order to strike a good balance between timbre disentanglement and sound quality, AutoVC requires imposing very strict constraints on the dimensionality of the latent representation. This constraint affects the quality of the converted speech while making it challenging to apply to other datasets directly. This paper proposes a new voice conversion framework that uses only one encoder to obtain timbre and content information by partitioning the latent space in the channel dimension. Furthermore, two different types of classifiers and two additional reconstruction losses are proposed to ensure that different parts of the latent space contain only separated content and timbre information, respectively. Experiments on the VCTK dataset show that the proposed model achieves state-of-the-art results in terms of the naturalness and similarity of converted speech. In addition, we experimentally show that for different division proportions of latent space, the content and timbre information will always be well separated.

Abstract Peptides and proteins are pharmacologically active and highly safe; however, their local delivery remains a challenge in the field of drug delivery. Based on this, a novel lysine‐based ionic liquid via proton transfer between lysine and malic acid that can overcome the difficulties in the application of copper peptides is developed. Density functional theory, molecular docking simulations, and experimental validation demonstrate the chemical stability of this ionic liquid. Furthermore, by combining it with a vegetable oil system, a biocompatible lysine‐based ionic liquid microemulsion system is constructed, which perfectly combines the high viscosity of ionic liquids and low viscosity of microemulsions. Multi‐dimensional analysis shows that the system not only leads to a dramatic elevation in the local penetration of copper peptides, but also enables copper peptides to effectively suppress skin aging. In addition, transcriptome sequencing analysis reveals that the system primarily regulates p53, p16, and ECM‐related signaling pathways to combat photoaging in the skin. Overall, a noninvasive, efficient, and convenient anti‐aging system with enormous potential is developed for the clinical prevention and treatment of skin photoaging.

Recovering the masked speech frames is widely applied in speech representation learning. However, most of these models use random masking in the pre-training. In this work, we proposed two kinds of masking approaches: (1) speech-level masking, making the model to mask more speech segments than silence segments, (2) phoneme-level masking, forcing the model to mask the whole frames of the phoneme, instead of phoneme pieces. We pre-trained the model via these two approaches, and evaluated on two downstream tasks, phoneme classification and speaker recognition. The experiments demonstrated that the proposed masking approaches are beneficial to improve the performance of speech representation.

Scene text recognition, especially irregular text recognition, is a challenging task due to the large variance in text appearance. Although some existing methods have achieved state-of-the-art performance with the attention-based encoder-decoder framework, they always perform poorly on some challenging text such as severely curved, blurred, and incomplete-semantic text. To address these issues, we propose a Dual-Branch Cross-Attention Network (DBCAN). Different from the previous methods heavily relying on semantic information, DBCAN can enhance the position clues and learn semantic relations with two separate branches and fuse them by a tailored Cross-Attention Module (CAM). Furthermore, a Convolution-Based 2D Positional Embedding (CBPE) is introduced to describe the 2D spatial dependencies of characters. Extensive experiments demonstrate our DBCAN is more accurate and robust than the previous methods and achieves state-of-the-art performance on several benchmarks, particularly CUTE (93.4%). Our code is made publicly available at https://github.com/GaoXinJian-USTC/DBCAN.

The ventrolateral prefrontal cortices (VLPFC) are crucial regions involved in voluntary emotion regulation. However, the lateralization of the VLPFC in downregulating negative emotions remains unclear; and whether the causal role of the VLPFC is generalizable to upregulating positive emotions is unexplored. This study used transcranial magnetic stimulation (TMS) to examine the causal relationship between the left/right VLPFC and social emotion reappraisal. One hundred and twenty participants were randomly assigned to either active (left and right VLPFC groups, n = 40/40) or sham (vertex, n = 40) TMS groups. Participants were instructed to passively receive social feedback or use reappraisal strategies to positively regulate their emotions. While the subjective emotional rating showed that the bilateral VLPFC facilitated the reappraisal success, the electrophysiological measure of the late positive potential (LPP) demonstrated a more critical role of the right VLPFC on social pain relief (decreased LPP amplitudes) and social reward magnification (enhanced LPP amplitudes). In addition, the influence of emotion regulation on social evaluation was found to be mediated by the memory of social feedback, indicating the importance of memory in social behavioral shaping. These findings suggest clinical protocols for the rehabilitation of emotion‐regulatory function in patients with affective and social disorders.

This paper introduces a dual learning system for neural voice conversion (DualVC) using relatively few samples based on the symmetry of the speech conversion task. The system contains a pair of sequence-to-sequence neural networks that have the same structure but are trained in opposite directions. The objective function of the dual model training is the sum of paired conversion loss and reconstruction loss during the dual training circle. The models in the two directions are trained alternately to guide each other by the corresponding reconstruction loss. Furthermore, curriculum learning techniques are used to load models in existing fields into the current task to accelerate the rapid iteration and convergence of the model. The experiment on the voice conversion task with the proposed DualVC and curriculum learning strategy obtained a comparable naturalness and similarity with only a 30% dataset than the BaseVC model trained on the full dataset.

Abstract Accurate organ at risk (OAR) segmentation is critical to reduce the radiotherapy post-treatment complications. Consensus guidelines recommend a set of more than 40 OARs in the head and neck (H&N) region, however, due to the predictable prohibitive labor-cost of this task, most institutions choose a substantially simplified protocol by delineating a smaller subset of OARs and neglecting the dose distributions associated with other OARs. In this work we propose a novel, automated and highly effective stratified OAR segmentation (SOARS) system using deep learning to precisely delineate a comprehensive set of 42 H&N OARs. SOARS stratifies 42 OARs into anchor, mid-level, and small & hard subcategories, with specifically derived neural network architectures for each category by neural architecture search (NAS) principles. We built SOARS models using 176 training patients in an internal institution and independently evaluated on 1327 external patients across six different institutions. It consistently outperformed other state-of-the-art methods by at least 3-5% in Dice score for each institutional evaluation (up to 36% relative error reduction in other metrics). More importantly, extensive multi-user studies evidently demonstrated that 98% of the SOARS predictions need only very minor or no revisions for direct clinical acceptance (saving 90% radiation oncologists workload), and their segmentation and dosimetric accuracy are within or smaller than the inter-user variation. These findings confirmed the strong clinical applicability of SOARS for the OAR delineation process in H&N cancer radiotherapy workflows, with improved efficiency, comprehensiveness, and quality.