Cross-modality magnetic resonance (MR) image synthesis aims to produce missing modalities from existing ones. Currently, several methods based on deep neural networks have been developed using both source- and target-modalities in a supervised learning manner. However, it remains challenging to obtain a large amount of completely paired multi-modal training data, which inhibits the effectiveness of existing methods. In this paper, we propose a novel Self-supervised Learning-based Multi-scale Transformer Network (SLMT-Net) for cross-modality MR image synthesis, consisting of two stages, \ie, a pre-training stage and a fine-tuning stage. During the pre-training stage, we propose an Edge-preserving Masked AutoEncoder (Edge-MAE), which preserves the contextual and edge information by simultaneously conducting the image reconstruction and the edge generation. Besides, a patch-wise loss is proposed to treat the input patches differently regarding their reconstruction difficulty, by measuring the difference between the reconstructed image and the ground-truth. In this case, our Edge-MAE can fully leverage a large amount of unpaired multi-modal data to learn effective feature representations. During the fine-tuning stage, we present a Multi-scale Transformer U-Net (MT-UNet) to synthesize the target-modality images, in which a Dual-scale Selective Fusion (DSF) module is proposed to fully integrate multi-scale features extracted from the encoder of the pre-trained Edge-MAE. Moreover, we use the pre-trained encoder as a feature consistency module to measure the difference between high-level features of the synthesized image and the ground truth one. Experimental results show the effectiveness of the proposed SLMT-Net, and our model can reliably synthesize high-quality images when the training set is partially unpaired. Our code will be publicly available at https://github.com/lyhkevin/SLMT-Net.

We study a novel and important communication pattern in large-scale model-parallel deep learning (DL), which we call cross-mesh resharding. This pattern emerges when the two paradigms of model parallelism - intra-operator and inter-operator parallelism - are combined to support large models on large clusters. In cross-mesh resharding, a sharded tensor needs to be sent from a source device mesh to a destination device mesh, on which the tensor may be distributed with the same or different layouts. We formalize this as a many-to-many multicast communication problem, and show that existing approaches either are sub-optimal or do not generalize to different network topologies or tensor layouts, which result from different model architectures and parallelism strategies. We then propose two contributions to address cross-mesh resharding: an efficient broadcast-based communication system, and an "overlapping-friendly" pipeline schedule. On microbenchmarks, our overall system outperforms existing ones by up to 10x across various tensor and mesh layouts. On end-to-end training of two large models, GPT-3 and U-Transformer, we improve throughput by 10% and 50%, respectively.

这里介绍了人工智能研究所（IARAI）组织的2022年Landslide4sense（L4S）竞赛的科学结果。竞争的目的是根据全球收集的卫星图像的大规模多个来源自动检测滑坡。 2022 L4S旨在促进有关使用卫星图像的语义分割任务的深度学习模型（DL）模型最新发展的跨学科研究。在过去的几年中，由于卷积神经网络（CNN）的发展，基于DL的模型已经达到了对图像解释的期望。本文的主要目的是介绍本次比赛中介绍的细节和表现最佳的算法。获胜的解决方案详细介绍了Swin Transformer，Segformer和U-NET等最先进的模型。还考虑了先进的机器学习技术和诸如硬采矿，自我培训和混合数据增强之类的策略。此外，我们描述了L4S基准数据集，以促进进一步的比较，并在线报告准确性评估的结果。可以在\ textIt {未来开发排行榜上访问数据，以供将来评估，\ url {https://www.iarai.ac.ac.at/landslide4sense/challenge/}，并邀请研究人员提交更多预测结果，评估准确性在他们的方法中，将它们与其他用户的方法进行比较，理想情况下，改善了本文报告的滑坡检测结果。

ALPA通过生成统一数据，操作员和管道并行性的执行计划来自动对大型深度学习（DL）模型的模型平行训练。现有的模型并行训练系统要求用户手动创建并行化计划，或者自动从有限的模型并行性配置中生成一个计划。它们不足以在分布式计算设备上扩展复杂的DL模型。 ALPA通过将并行性视为两个层次级别来分配大型DL模型的训练：操作员和操作员并行性。基于它，ALPA构建了一个新的分层空间，用于大规模的模型并行执行计划。 ALPA设计了许多汇编，以在每个并行性级别自动得出有效的并行执行计划。 ALPA实现了有效的运行时，以在分布式计算设备上协调两级并行执行。我们的评估表明，ALPA生成的并行化计划，即使在其设计的型号上，也可以匹配或超过手动模型并联训练系统。与专业系统不同，ALPA还推广到具有异质体系结构和模型的模型，而没有手动设计的计划。 ALPA的源代码可在https://github.com/alpa-projects/alpa上公开获得

Surgical robot automation has attracted increasing research interest over the past decade, expecting its huge potential to benefit surgeons, nurses and patients. Recently, the learning paradigm of embodied AI has demonstrated promising ability to learn good control policies for various complex tasks, where embodied AI simulators play an essential role to facilitate relevant researchers. However, existing open-sourced simulators for surgical robot are still not sufficiently supporting human interactions through physical input devices, which further limits effective investigations on how human demonstrations would affect policy learning. In this paper, we study human-in-the-loop embodied intelligence with a new interactive simulation platform for surgical robot learning. Specifically, we establish our platform based on our previously released SurRoL simulator with several new features co-developed to allow high-quality human interaction via an input device. With these, we further propose to collect human demonstrations and imitate the action patterns to achieve more effective policy learning. We showcase the improvement of our simulation environment with the designed new features and tasks, and validate state-of-the-art reinforcement learning algorithms using the interactive environment. Promising results are obtained, with which we hope to pave the way for future research on surgical embodied intelligence. Our platform is released and will be continuously updated in the website: https://med-air.github.io/SurRoL/

Recent advances in artificial intelligence (AI) have significantly intensified research in the geoscience and remote sensing (RS) field. AI algorithms, especially deep learning-based ones, have been developed and applied widely to RS data analysis. The successful application of AI covers almost all aspects of Earth observation (EO) missions, from low-level vision tasks like super-resolution, denoising, and inpainting, to high-level vision tasks like scene classification, object detection, and semantic segmentation. While AI techniques enable researchers to observe and understand the Earth more accurately, the vulnerability and uncertainty of AI models deserve further attention, considering that many geoscience and RS tasks are highly safety-critical. This paper reviews the current development of AI security in the geoscience and RS field, covering the following five important aspects: adversarial attack, backdoor attack, federated learning, uncertainty, and explainability. Moreover, the potential opportunities and trends are discussed to provide insights for future research. To the best of the authors' knowledge, this paper is the first attempt to provide a systematic review of AI security-related research in the geoscience and RS community. Available code and datasets are also listed in the paper to move this vibrant field of research forward.

基于文本描述的高分辨率遥感图像的合成在许多实际应用方案中具有巨大的潜力。尽管深度神经网络在许多重要的遥感任务中取得了巨大的成功，但是从文本描述中生成现实的遥感图像仍然非常困难。为了应对这一挑战，我们提出了一个新颖的文本形象现代霍普菲尔德网络（TXT2IMG-MHN）。 TXT2IMG-MHN的主要思想是在具有现代Hopfield层的文本和图像嵌入方式上进行层次原型学习。 TXT2IMG-MHN并没有直接学习具体但高度多样化的文本图像联合特征表示，而是旨在从文本图像嵌入中学习最具代表性的原型，从而实现一种粗略的学习策略。然后可以利用这些学到的原型来代表文本到图像生成任务中更复杂的语义。为了更好地评估生成图像的现实主义和语义一致性，我们使用对合成图像训练的分类模型对真实遥感数据进行零击分类。尽管它很简单，但我们发现，零弹性分类的总体准确性可以作为评估从文本生成图像的能力的良好指标。基准遥感文本图像数据集上的广泛实验表明，所提出的TXT2IMG-MHN比现有方法可以生成更现实的遥感图像。代码和预培训模型可在线获得（https://github.com/yonghaoxu/txt2img-mhn）。

求解部分微分方程（PDE）是物理，生物学和化学领域的重要研究手段。作为数值方法的近似替代方法，Pinn受到了广泛的关注，并在许多领域发挥了重要作用。但是，Pinn使用完全连接的网络作为其模型，在时间和空间中，其合适能力和有限的外推能力有限。在本文中，我们提出了用于求解图形神经网络基础的部分微分方程的phygnnet，该方程由编码器，处理器和解码器块组成。特别是，我们将计算区域划分为常规网格，在网格上定义部分差分运算符，然后构建PDE损失以使网络优化以构建Phygnnet模型。更重要的是，我们对汉堡方程和热方程式进行比较实验以验证我们的方法，结果表明，与PINN相比，我们的方法在时间和空间区域具有更好的拟合能力和外推能力。

计算机辅助的微创手术在使现代经营剧院受益方面具有巨大的潜力。从内窥镜流传输的视频数据提供了丰富的信息，以支持下一代智能手术系统的上下文意识。为了在手术过程中获得准确的感知和自动操纵，基于学习的技术是一种有希望的方法，近年来可以实现先进的图像分析和场景理解。但是，学习此类模型高度依赖于大规模，高质量和多任务标签的数据。目前，这是该主题的瓶颈，因为可用的公共数据集在CAI领域仍然非常有限。在本文中，我们介绍并发布了第一个具有多个基于图像的感知任务的集成数据集（称为Autolaparo），以促进子宫切除术手术中的基于学习的自动化。我们的Autolaparo数据集是根据整个子宫切除术程序的全长视频开发的。具体而言，数据集中制定了三个不同但高度相关的任务，包括手术工作流识别，腹腔镜运动预测以及仪器和关键解剖学细分。此外，我们还提供了最先进模型的实验结果，作为参考基准，用于该数据集的进一步模型开发和评估。该数据集可从https://autolaparo.github.io获得。

无锚的检测器基本上将对象检测作为密集的分类和回归。对于流行的无锚检测器，通常是引入单个预测分支来估计本地化的质量。当我们深入研究分类和质量估计的实践时，会观察到以下不一致之处。首先，对于某些分配了完全不同标签的相邻样品，训练有素的模型将产生相似的分类分数。这违反了训练目标并导致绩效退化。其次，发现检测到具有较高信心的边界框与相应的地面真相具有较小的重叠。准确的局部边界框将被非最大抑制（NMS）过程中的精确量抑制。为了解决不一致问题，提出了动态平滑标签分配（DSLA）方法。基于最初在FCO中开发的中心概念，提出了平稳的分配策略。在[0，1]中将标签平滑至连续值，以在正样品和负样品之间稳定过渡。联合（IOU）在训练过程中会动态预测，并与平滑标签结合。分配动态平滑标签以监督分类分支。在这样的监督下，质量估计分支自然合并为分类分支，这简化了无锚探测器的体系结构。全面的实验是在MS Coco基准上进行的。已经证明，DSLA可以通过减轻上述无锚固探测器的不一致来显着提高检测准确性。我们的代码在https://github.com/yonghaohe/dsla上发布。