Jianghao Lin, Yuanyuan Shi, Xin Peng, Renjie Ding, Hairui Wang, Yuxuan Peng, Bizhe Bai, Weixi Song, Fengshuo Bai, Huacan Chai, Weinan Zhang, Fei Huang, Ying Wen. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2026.

Surgical scene understanding is a vital intelligent technique in robot-assisted surgery, including surgical instrument detection, segmentation, and instrument–tissue interaction detection. Existing methods typically address these tasks in isolation, neglecting the intrinsic correlations among them. In this work, we innovatively propose a unified multitask framework named UniSurg, being the first to jointly address these three critical aspects of surgical scene understanding, thereby providing the robot with multidimensional perceptual capabilities. By exploring the inter-task correlations and reusing shared features, UniSurg has been demonstrated to significantly enhance the scene analysis performance. To address pose variability of the instruments under the constrained field of view in laparoscopic surgery, we design an Attention Enhanced Conditional Convolution (AEC-Conv) that dynamically adjusts kernels based on pose-specific features for improved adaptability. To further enhance interaction detection, we propose the Temporal Difference Enhancement module (TDE), which captures motion cues by amplifying inter-frame differences, and the Pyramid Global Feature Enhancement module (PGFE), which leverages graph-based hierarchical context to model global relational dependencies. Experiments on the Endovis2018 dataset and a clinical multitask dataset MILVis demonstrate the superior multitask performance of UniSurg.

High-precision robotic assembly remains a challenge in intelligent manufacturing. Existing vision-based approaches often learn predefined trajectories from large image corpora yet underutilize task-relevant visual cues during execution, limiting deployment. We present a diffusion-based end-to-end assembly policy that performs redundancy-aware cross-scale fusion and provides stage-dependent conditioning for action generation. Specifically, we introduce a bidirectional-attention complementary visual fusion module that aligns cross-scale observations and produces scale-consistent features with reduced redundancy. We then introduce a dual-stream adaptive modulation module that enables time-state correlated routing and progressively shifts emphasis from scene-level stabilization to contact-level refinement within the denoising process. Coupling complementary visual fusion with adaptive-modulation-conditioned denoising, we develop a staged modulation diffusion policy for end-to-end action generation, producing temporally coherent and geometrically accurate actions. Experiments on five real-world tasks demonstrate consistent improvements over representative baselines, with average gains of 41.10% in overall task success rate and 63.16% in precise assembly rate on three representative assembly benchmarks.

Unsupervised anomaly detection (UAD) methods typically detect anomalies by learning and reconstructing the normative distribution. However, since anomalies constantly invade and affect their surroundings, sub-healthy areas in the junction present structural deformations that could be easily misidentified as anomalies, posing difficulties for UAD methods that solely learn the normative distribution. The use of multimodal images can facilitate to address the above challenges, as they can provide complementary information of anomalies. Therefore, this paper propose a novel method for UAD in preoperative multimodal images, called Erasure Perception Diffusion model (EPDiff). First, the Local Erasure Progressive Training (LEPT) framework is designed to better rebuild sub-healthy structures around anomalies through the diffusion model with a two-phase process. Initially, healthy images are used to capture deviation features labeled as potential anomalies. Then, these anomalies are locally erased in multimodal images to progressively learn sub-healthy structures, obtaining a more detailed reconstruction around anomalies. Second, the Global Structural Perception (GSP) module is developed in the diffusion model to realize global structural representation and correlation within images and between modalities through interactions of high-level semantic information. In addition, a training-free module, named Multimodal Attention Fusion (MAF) module, is presented for weighted fusion of anomaly maps between different modalities and obtaining binary anomaly outputs. Experimental results show that EPDiff improves the AUPRC and mDice scores by 2% and 3.9% on BraTS2021, and by 5.2% and 4.5% on Shifts over the state-of-the-art methods, which proves the applicability of EPDiff in diverse anomaly diagnosis. The code is available at https://github.com/wjiazheng/EPDiff.

As a key enabling technology in intelligent heterogeneous Internet of Things (IoT), edge caching provides important support for reducing core network load and improving network service efficiency, especially for high bandwidth demand services represented by multimedia applications. However, external time-varying information is hard to be obtained comprehensively in a complicated heterogeneous IoT environment. Meanwhile, there exists the substitutability of content (e.g., videos with different bitrates), which is difficult to make caching decisions online in real-time to achieve fast feedback with low latency and avoid useless deployment. To this end, this article designs a transcoding-enabled online cache scheme for IoT video service with cloud–edge–terminal collaboration. First, we design a variable bitrate video routing strategy to dynamically retrieve content from cloud/edge according to user demands. Furthermore, the video caching problem is considered as an online convex optimization problem to learn utility gradient and determine the optimal caching strategy in real-time without any prior information. On this basis, we extend the problem to elastic networks with dynamic available resources and prove the sublinear regret and sublinear constraint violation. Finally, we summarized five video request data sets and carried out differentiated multiple verifications based on different request habits and content requirements. Compared with the most advanced algorithms in terms of delay, we evaluated the performance advantages of the proposed scheme.

Surgical instrument segmentation is of great significance to robot-assisted surgery, but the noise caused by reflection, water mist, and motion blur during the surgery as well as the different forms of surgical instruments would greatly increase the difficulty of precise segmentation. A novel method called Branch Aggregation Attention network (BAANet) is proposed to address these challenges, which adopts a lightweight encoder and two designed modules, named Branch Balance Aggregation module (BBA) and Block Attention Fusion module (BAF), for efficient feature localization and denoising. By introducing the unique BBA module, features from multiple branches are balanced and optimized through a combination of addition and multiplication to complement strengths and effectively suppress noise. Furthermore, to fully integrate the contextual information and capture the region of interest, the BAF module is proposed in the decoder, which receives adjacent feature maps from the BBA module and localizes the surgical instruments from both global and local perspectives by utilizing a dual branch attention mechanism. According to the experimental results, the proposed method has the advantage of being lightweight while outperforming the second-best method by 4.03%, 1.53%, and 1.34% in mIoU scores on three challenging surgical instrument datasets, respectively, compared to the existing state-of-the-art methods. Code is available at https://github.com/SWT-1014/BAANet.

Virtual reality (VR) content, including 360° panoramic video, provides users with an immersive multimedia experience and therefore attracts increasing research and development attention. However, the requirement of high bandwidth and low latency of virtual reality service demand puts forward greater challenges to the current infrastructure, especially mobile networks. Inspired by the sharable nature of virtual reality content tiles, we further considered the potential opportunities for computing, caching, and multicast to address the challenges of transmission of panoramic content. This paper proposes a novel transcoding-enabled VR video caching and delivery framework for edge-enhanced next-generation wireless networks. Firstly, an edge cooperative caching scheme based on multi-agent reinforcement learning is introduced to improve the utilization efficiency of computing and storage resources, and then reduce service delay. Second, a two-tier NOMA-based base station-multicast group matching mechanism is designed to solve the collaboration challenge during the edge delivery process. A series of experiments have demonstrated the advantages of the proposed scheme in terms of cache hit rate, latency and other aspects in comparison with alternative approaches.

This paper proposes a novel neural network approach for low frequency oscillation mode identification in power systems. After employing the fast Fourier transform in order selection, we then introduce a specific neural network whose topology strictly follows the exponentially damped sinusoidal model of low frequency oscillation signals. We thus turn the parameter estimation into an optimization problem. Simulations show that the proposed approach is superior to the state-of-art neural network schemes in anti-noise ability, parameter accuracy and computation speed.