The lack of efficient segmentation methods and fully-labeled datasets limits the comprehensive assessment of optical coherence tomography angiography (OCTA) microstructures like retinal vessel network (RVN) and foveal avascular zone (FAZ), which are of great value in ophthalmic and systematic diseases evaluation. Here, we introduce an innovative OCTA microstructure segmentation network (OMSN) by combining an encoder-decoder-based architecture with multi-scale skip connections and the split-attention-based residual network ResNeSt, paying specific attention to OCTA microstructural features while facilitating better model convergence and feature representations. The proposed OMSN achieves excellent single/multi-task performances for RVN or/and FAZ segmentation. Especially, the evaluation metrics on multi-task models outperform single-task models on the same dataset. On this basis, a fully annotated retinal OCTA segmentation (FAROS) dataset is constructed semi-automatically, filling the vacancy of a pixel-level fully-labeled OCTA dataset. OMSN multi-task segmentation model retrained with FAROS further certifies its outstanding accuracy for simultaneous RVN and FAZ segmentation.

A learned system uses machine learning (ML) internally to improve performance. We can expect such systems to be vulnerable to some adversarial-ML attacks. Often, the learned component is shared between mutually-distrusting users or processes, much like microarchitectural resources such as caches, potentially giving rise to highly-realistic attacker models. However, compared to attacks on other ML-based systems, attackers face a level of indirection as they cannot interact directly with the learned model. Additionally, the difference between the attack surface of learned and non-learned versions of the same system is often subtle. These factors obfuscate the de-facto risks that the incorporation of ML carries. We analyze the root causes of potentially-increased attack surface in learned systems and develop a framework for identifying vulnerabilities that stem from the use of ML. We apply our framework to a broad set of learned systems under active development. To empirically validate the many vulnerabilities surfaced by our framework, we choose 3 of them and implement and evaluate exploits against prominent learned-system instances. We show that the use of ML caused leakage of past queries in a database, enabled a poisoning attack that causes exponential memory blowup in an index structure and crashes it in seconds, and enabled index users to snoop on each others' key distributions by timing queries over their own keys. We find that adversarial ML is a universal threat against learned systems, point to open research gaps in our understanding of learned-systems security, and conclude by discussing mitigations, while noting that data leakage is inherent in systems whose learned component is shared between multiple parties.

Conceptual knowledge is fundamental to human cognition and knowledge bases. However, existing knowledge probing works only focus on evaluating factual knowledge of pre-trained language models (PLMs) and ignore conceptual knowledge. Since conceptual knowledge often appears as implicit commonsense behind texts, designing probes for conceptual knowledge is hard. Inspired by knowledge representation schemata, we comprehensively evaluate conceptual knowledge of PLMs by designing three tasks to probe whether PLMs organize entities by conceptual similarities, learn conceptual properties, and conceptualize entities in contexts, respectively. For the tasks, we collect and annotate 24k data instances covering 393 concepts, which is COPEN, a COnceptual knowledge Probing bENchmark. Extensive experiments on different sizes and types of PLMs show that existing PLMs systematically lack conceptual knowledge and suffer from various spurious correlations. We believe this is a critical bottleneck for realizing human-like cognition in PLMs. COPEN and our codes are publicly released at https://github.com/THU-KEG/COPEN.

The formalization of existing mathematical proofs is a notoriously difficult process. Despite decades of research on automation and proof assistants, writing formal proofs remains arduous and only accessible to a few experts. While previous studies to automate formalization focused on powerful search algorithms, no attempts were made to take advantage of available informal proofs. In this work, we introduce Draft, Sketch, and Prove (DSP), a method that maps informal proofs to formal proof sketches, and uses the sketches to guide an automated prover by directing its search to easier sub-problems. We investigate two relevant setups where informal proofs are either written by humans or generated by a language model. Our experiments and ablation studies show that large language models are able to produce well-structured formal sketches that follow the same reasoning steps as the informal proofs. Guiding an automated prover with these sketches enhances its performance from 20.9% to 39.3% on a collection of mathematical competition problems.

基于文本的视觉问题回答〜（TextVQA）旨在为具有多个场景文本的图像问题提供正确的答案。在大多数情况下，文本自然附着在物体表面上。因此，文本和对象之间的空间推理在文本VQA中至关重要。但是，现有方法在从输入图像中学到的2D空间信息中受到限制，并依靠基于变压器的体系结构在融合过程中隐含地推理。在此设置下，这些2D空间推理方法无法区分同一图像平面上的视觉对象和场景文本之间的细颗粒空间关系，从而损害了TextVQA模型的可解释性和性能。在本文中，我们将3D几何信息引入了类似人类的空间推理过程，以逐步捕获关键对象的上下文知识。 ％我们通过引入3D几何信息来捕获关键对象的上下文知识来制定类似人类的空间推理过程。为了增强模型对3D空间关系的理解，特别是（i）〜我们提出了一个关系预测模块，以准确定位关键对象的关注区域； （ii）〜我们设计了一个深度感知的注意校准模块，以根据关键对象校准OCR令牌的注意力。广泛的实验表明，我们的方法在TextVQA和ST-VQA数据集上实现了最先进的性能。更令人鼓舞的是，我们的模型在涉及TextVQA和ST-VQA有效拆分中的空间推理的问题上以5.7 \％和12.1 \％的明显边缘超过了他人。此外，我们还验证了模型对基于文本的图像字幕任务的普遍性。

本文回顾了AIM 2022上压缩图像和视频超级分辨率的挑战。这项挑战包括两条曲目。轨道1的目标是压缩图像的超分辨率，轨迹〜2靶向压缩视频的超分辨率。在轨道1中，我们使用流行的数据集DIV2K作为培训，验证和测试集。在轨道2中，我们提出了LDV 3.0数据集，其中包含365个视频，包括LDV 2.0数据集（335个视频）和30个其他视频。在这一挑战中，有12支球队和2支球队分别提交了赛道1和赛道2的最终结果。所提出的方法和解决方案衡量了压缩图像和视频上超分辨率的最先进。提出的LDV 3.0数据集可在https://github.com/renyang-home/ldv_dataset上找到。此挑战的首页是在https://github.com/renyang-home/aim22_compresssr。

几十年来，手写的中文文本识别（HCTR）一直是一个活跃的研究主题。但是，大多数以前的研究仅关注裁剪文本图像的识别，而忽略了实际应用程序中文本线检测引起的错误。尽管近年来已经提出了一些针对页面文本识别的方法，但它们要么仅限于简单布局，要么需要非常详细的注释，包括昂贵的线条级别甚至角色级边界框。为此，我们建议Pagenet端到端弱监督的页面级HCTR。 Pagenet检测并识别角色并预测其之间的阅读顺序，在处理复杂的布局（包括多方向和弯曲的文本线路）时，这更健壮和灵活。利用所提出的弱监督学习框架，Pagenet只需要对真实数据进行注释。但是，它仍然可以在字符和线级别上输出检测和识别结果，从而避免标记字符和文本线条的界限框的劳动和成本。在五个数据集上进行的广泛实验证明了Pagenet优于现有的弱监督和完全监督的页面级方法。这些实验结果可能会引发进一步的研究，而不是基于连接主义时间分类或注意力的现有方法的领域。源代码可在https://github.com/shannanyinxiang/pagenet上获得。

在线和离线手写的中文文本识别（HTCR）已经研究了数十年。早期方法采用了基于过度裂段的策略，但遭受低速，准确性不足和角色分割注释的高成本。最近，基于连接主义者时间分类（CTC）和注意机制的无分割方法主导了HCTR的领域。但是，人们实际上是按字符读取文本的，尤其是对于中文等意识形态图。这就提出了一个问题：无细分策略真的是HCTR的最佳解决方案吗？为了探索此问题，我们提出了一种基于细分的新方法，用于识别使用简单但有效的完全卷积网络实现的手写中文文本。提出了一种新型的弱监督学习方法，以使网络仅使用笔录注释进行训练。因此，可以避免以前基于细分的方法所需的昂贵字符分割注释。由于缺乏完全卷积网络中的上下文建模，我们提出了一种上下文正则化方法，以在培训阶段将上下文信息集成到网络中，这可以进一步改善识别性能。在四个广泛使用的基准测试中进行的广泛实验，即Casia-HWDB，Casia-Olhwdb，ICDAR2013和Scut-HCCDOC，表明我们的方法在线和离线HCTR上都显着超过了现有方法，并且表现出比CTC/ CTC/ CTC/ CTC/ CTC/速度高得多的方法。基于注意力的方法。

反转技术被广泛用于重建基于表面的地球物理测量值（例如，地震，电气/磁（EM）数据）的地下物理特性（例如，速度，电导率）。这些问题受波浪或麦克斯韦方程等部分微分方程（PDE）的控制。解决地球物理反演问题由于不适当和高计算成本而具有挑战性。为了减轻这些问题，最近的研究利用深层神经网络来学习从测量到物业的倒置映射。在本文中，我们表明，这样的映射可以通过仅有五层的非常浅（但不是宽）网络来很好地建模。这是基于我们对有趣属性的新发现来实现的：在高维空间中应用积分变换后，输入和输出之间的近乎线性关系。特别是，在处理由波方程控制的从地震数据到地下速度的反演时，与高斯核的速度的积分结果与正弦核的地震数据的积分线性相关。此外，该属性可以轻松地转变为用于反转的轻质编码器网络。编码器包含地震数据和线性转换的整合，而无需进行微调。解码器仅由一个单个变压器块组成，以逆转速度的积分。实验表明，这种有趣的属性可用于四个不同数据集的两个地球物理倒置问题。与更深的倒置网络相比，我们的方法达到了可比的精度，但消耗的参数大大减少。

本文报告了Chalearn的Autodl挑战系列的结果和后攻击分析，这有助于对自动学习（DL）进行分类，以便在各种环境中引入的深度学习（DL），但缺乏公平的比较。格式化所有输入数据模型（时间序列，图像，视频，文本，表格）作为张量，所有任务都是多标签分类问题。代码提交已在隐藏的任务上执行，具有限制时间和计算资源，推动快速获取结果的解决方案。在此设置中，DL方法占主导地位，但流行的神经结构搜索（NAS）是不切实际的。解决方案依赖于微调预培训的网络，架构匹配数据模块。挑战后测试没有透露超出强加时间限制的改进。虽然没有组件尤其原始或新颖，但是一个高级模块化组织出现了“Meta-Learner”，“数据摄入”，“模型选择器”，“模型/学习者”和“评估员”。这种模块化使得消融研究，揭示了（离坡）元学习，合奏和高效数据管理的重要性。异构模块组合的实验进一步证实了获胜解决方案的（本地）最优性。我们的挑战队遗产包括一个持久的基准（http://utodl.chalearn.org），获胜者的开放源代码，以及免费的“autodl自助服务”。