Image fusion aims to synthesize a single high-quality image from a pair of inputs captured under challenging conditions, such as differing exposure levels or focal depths. A core challenge lies in effectively handling disparities in dynamic range and focus depth between the inputs. With the advent of vision–language models, recent methods incorporate textual descriptions as auxiliary guidance to enhance fusion quality. However, simply incorporating coarse-grained descriptions hampers the understanding of fine-grained details and poses challenges for precise cross-modal alignment. To address these limitations, we propose Multi-grained Text-guided Image Fusion (MTIF), a novel fusion paradigm with three key designs. First, it introduces multi-grained textual descriptions that separately capture fine details, structural cues, and semantic content, guiding image fusion through a hierarchical cross-modal modulation module. Second, it involves supervision signals at each granularity to facilitate alignment between visual and textual features and enhance the utility of auxiliary text. Third, it adopts a saliency-driven enrichment module to augment training data with dense semantic content, further strengthening the cross-modal modulation and alignment. Extensive experiments show that MTIF consistently outperforms previous methods on both multi-exposure and multi-focus image fusion tasks.

ABSTRACT Aspect‐Based Sentiment Triplet Extraction (ASTE) is one of the hot topics in recent years. Relevant researchers have proposed many neural network models for aspect‐based sentiment triplet extraction. However, they fail to model the complex syntactic and semantic associations between tokens, which limits the synergy of features from different angles, which may lead to inaccurate encoding of the relationship between tokens and thus inaccurate extraction of triples. In response to the problems mentioned above, a graph attention aggregation network with multi‐level Syntactic and Semantic Enhanced Prompts for aspect‐Based sentiment triple extraction (SSEP) is proposed. First, a graph attention relation aggregation module is designed in the context encoding part. Specifically, the module first constructs a relation aggregation graph through the output of the pre‐trained language model, then designs a graph attention aggregator, and finally aggregates the multi‐level output of the pre‐trained language model according to the relation aggregation graph and the graph attention aggregator. Second, a syntactic and semantic enhanced prompt module is proposed. The module uses relation table attention to prompt the model, and then uses a dual‐channel graph neural network to further enhance syntactic and semantic information and prune unimportant information. In addition, a joint boundary detection module is designed. The module can directly extract sentiment triplets using joint boundary detection labels. Finally, experimental results on four public datasets show that SSEP achieves state‐of‐the‐art performance and outperforms other models.

The integration of hardware accelerators has significantly advanced the capabilities of modern recommendation systems, enabling the exploration of complex ranking paradigms previously deemed impractical. However, the GPU-based computational costs present substantial challenges. In this paper, we demonstrate our development of an efficiency-driven approach to explore these paradigms, moving beyond traditional reliance on native PyTorch modules. We address the specific challenges posed by ranking models’ dependence on categorical features, which vary in length and complicate GPU utilization. We introduce Jagged Feature Interaction Kernels, a novel method designed to extract fine-grained insights from long categorical features through efficient handling of dynamically sized tensors. We further enhance the performance of attention mechanisms by integrating Jagged tensors with Flash Attention. Our novel Jagged Flash Attention achieves up to 9 × speedup and 22 × memory reduction compared to dense attention. Notably, it also outperforms dense flash attention, with up to 3 × speedup and 53% more memory efficiency. In production models, we observe 10% QPS improvement and 18% memory savings, enabling us to scale our recommendation systems with longer features and more complex architectures.

Abstract Aspect-based sentiment analysis aims to identify the sentiment polarity of aspects in a given sentence. Although existing neural network models show promising results, they cannot meet the expectations in the case of a single network structure and limited dataset. When an aspect term composes more than one word, many models use the coarse-grained attention mechanism but lead to the unsatisfactory results. Besides, the relative distance between words in a sentence is always out of consideration. In this paper, we propose a model based on the interaction matrix and global attention mechanism to improve the ability of aspect-based sentiment analysis. First of all, the relative distance features of words in a sentence are initialized to enrich word embedding. Second, classic neural networks are applied to extract the essential features of word embedding in a sentence, such as long short-term memory and convolutional neural network. Third, an interaction matrix and global attention mechanism are combined to calculate weighted scores and measure relationships between aspect terms and context words. Finally, sentiment polarity is represented through a softmax layer. Experimental results on restaurant, laptop and twitter datasets show that the performance of the proposed model is superior to other methods.

It is an obvious fact that drivers’ drowsiness is more likely to cause traffic accidents. Recently, driver drowsiness detection has drawn considerable attention. In this paper, a novel drowsiness detection scheme is proposed, which can recognize drivers’ drowsiness actions through their facial expressions. First, a drowsiness action recognition model based on 3D-CNN is proposed, which can effectively distinguish drivers’ drowsiness actions and nondrowsiness actions. Second, a fusion algorithm of the two input streams is proposed, which can fuse gray image sequence and optical image sequence containing target motion information. Finally, the proposed model is evaluated on National Tsinghua University Driver Drowsiness Detection (NTHU-DDD) dataset. The experimental results show that the algorithm performs better than other algorithms, and its accuracy reaches 86.64%.

Knowledge graph has become an essential tool for semantic analysis with the development of natural language processing and deep learning. A high-quality knowledge graph is handy for building a high-performance knowledge-driven application. Despite recent advances in information extraction (IE) techniques, no suitable automated methods can be applied to constructing a domain-specific, comprehensive, and high-quality knowledge graph. However, a semi-automatic strategy, which can ensure the basic quality requirements of a knowledge graph, has been successfully implemented in the elementary science domain. This paper presents a semantic annotation system developed for building a high-quality legal knowledge graph (SALKG) using the semi-automatic strategy. We introduce its system design, architecture, algorithms, functions, and implementation. To investigate the effectiveness of SALKG, we conduct a preliminary annotation experiment with 280 legal texts which were collected from the Harvard Caselaw Access Project. The user evaluation from 32 graduate students demonstrates the high usability of SALKG in semantic annotation and the potential for building a high-quality legal knowledge graph. The system can also be adapted to other fields for constructing domain-specific knowledge graphs.

In this paper, resource allocation problem for multicast service in OFDM-based wireless cellular networks is investigated and a coding-based opportunistic scheduling scheme is proposed. In the proposed scheme, according to frequency resource and time resource, layered coding and erasure-correction coding are adopted, respectively. Layered coding divides the data into base layer and enhancement layer, which transmits the data in different rates to maximize the system capacity; erasure-correction coding compensates for the possible packet loss due to instantaneous channel information. To complete subcarrier allocation and bit loading for the coding-based scheme, a practical algorithm is designed. As only the users' mean channel gains are required, the proposed scheme reduces the overheads required for instantaneous channel information and is suitable for a fast time-varying fading environment. Numerical results show that the coding-based scheme can get significant performance improvement compared with the scheme without coding.

In this paper, we investigate resource allocation issue in OFDM-based wireless multicast system and joint subcarrier and power allocation scheme is proposed. The optimal solution of the combinable allocation has high computational complexity, so the solution is divided into two steps. In the first step, the subcarriers are allocated to the multicast services, and all the users that have subscribed to one service share the same subcarriers. In order to increase the system capacity, a subcarrier cooperation and user grouping method is proposed. To further lower the computational complexity, a simplified method is derived. In the second step, the water-filling method is adopted to get further capacity improvement. The simulation results show that the proposed scheme performs better than the existing schemes, and the proposed simplified scheme can get a similar performance. Besides, the resource allocation schemes with multi-dimension water-filling allocation notably outperform the schemes with equal power allocation.

In this paper, we investigate resource allocation issue in OFDMA-based decode-and-forward cooperative networks and propose joint subcarrier and power allocation schemes. The optimal solution of this combinable allocation has high computational complexity, so we divide our solution into two steps. The first step is to distribute subcarriers to relays and destination under the assumption of equal power distribution. Here, we propose Proportional Allocation (PA) strategy to achieve tradeoff between total throughput and fairness. To further improve the system performance, we introduce threshold into PA strategy, named Proportional Allocation with Threshold (PA-T), where subcarriers with bad performance are prevented from transmitting. Next, water-filling method is adopted to distribute the power to cooperative links in order to fully utilize the limited power. Simulation results show that system performance of the proposed schemes is significantly enhanced compared with an existing resource allocation scheme. Besides, the resource allocation schemes with water-filling method notably outperform schemes with equal power allocation.