State-of-the-art computer vision systems are trained to predict a fixed set of predetermined object categories. This restricted form of supervision limits their generality and usability since additional labeled data is needed to specify any other visual concept. Learning directly from raw text about images is a promising alternative which leverages a much broader source of supervision. We demonstrate that the simple pre-training task of predicting which caption goes with which image is an efficient and scalable way to learn SOTA image representations from scratch on a dataset of 400 million (image, text) pairs collected from the internet. After pre-training, natural language is used to reference learned visual concepts (or describe new ones) enabling zero-shot transfer of the model to downstream tasks. We study the performance of this approach by benchmarking on over 30 different existing computer vision datasets, spanning tasks such as OCR, action recognition in videos, geo-localization, and many types of fine-grained object classification. The model transfers non-trivially to most tasks and is often competitive with a fully supervised baseline without the need for any dataset specific training. For instance, we match the accuracy of the original ResNet-50 on ImageNet zero-shot without needing to use any of the 1.28 million training examples it was trained on. We release our code and pre-trained model weights at https://github.com/OpenAI/CLIP.

Implicit Neural Representations (INR) have recently shown to be powerful tool for high-quality video compression. However, existing works are limiting as they do not explicitly exploit the temporal redundancy in videos, leading to a long encoding time. Additionally, these methods have fixed architectures which do not scale to longer videos or higher resolutions. To address these issues, we propose NIRVANA, which treats videos as groups of frames and fits separate networks to each group performing patch-wise prediction. This design shares computation within each group, in the spatial and temporal dimensions, resulting in reduced encoding time of the video. The video representation is modeled autoregressively, with networks fit on a current group initialized using weights from the previous group's model. To further enhance efficiency, we perform quantization of the network parameters during training, requiring no post-hoc pruning or quantization. When compared with previous works on the benchmark UVG dataset, NIRVANA improves encoding quality from 37.36 to 37.70 (in terms of PSNR) and the encoding speed by 12X, while maintaining the same compression rate. In contrast to prior video INR works which struggle with larger resolution and longer videos, we show that our algorithm is highly flexible and scales naturally due to its patch-wise and autoregressive designs. Moreover, our method achieves variable bitrate compression by adapting to videos with varying inter-frame motion. NIRVANA achieves 6X decoding speed and scales well with more GPUs, making it practical for various deployment scenarios.

Supervised approaches generally rely on majority-based labels. However, it is hard to achieve high agreement among annotators in subjective tasks such as hate speech detection. Existing neural network models principally regard labels as categorical variables, while ignoring the semantic information in diverse label texts. In this paper, we propose AnnoBERT, a first-of-its-kind architecture integrating annotator characteristics and label text with a transformer-based model to detect hate speech, with unique representations based on each annotator's characteristics via Collaborative Topic Regression (CTR) and integrate label text to enrich textual representations. During training, the model associates annotators with their label choices given a piece of text; during evaluation, when label information is not available, the model predicts the aggregated label given by the participating annotators by utilising the learnt association. The proposed approach displayed an advantage in detecting hate speech, especially in the minority class and edge cases with annotator disagreement. Improvement in the overall performance is the largest when the dataset is more label-imbalanced, suggesting its practical value in identifying real-world hate speech, as the volume of hate speech in-the-wild is extremely small on social media, when compared with normal (non-hate) speech. Through ablation studies, we show the relative contributions of annotator embeddings and label text to the model performance, and tested a range of alternative annotator embeddings and label text combinations.

360-degree panoramic videos have gained considerable attention in recent years due to the rapid development of head-mounted displays (HMDs) and panoramic cameras. One major problem in streaming panoramic videos is that panoramic videos are much larger in size compared to traditional ones. Moreover, the user devices are often in a wireless environment, with limited battery, computation power, and bandwidth. To reduce resource consumption, researchers have proposed ways to predict the users' viewports so that only part of the entire video needs to be transmitted from the server. However, the robustness of such prediction approaches has been overlooked in the literature: it is usually assumed that only a few models, pre-trained on past users' experiences, are applied for prediction to all users. We observe that those pre-trained models can perform poorly for some users because they might have drastically different behaviors from the majority, and the pre-trained models cannot capture the features in unseen videos. In this work, we propose a novel meta learning based viewport prediction paradigm to alleviate the worst prediction performance and ensure the robustness of viewport prediction. This paradigm uses two machine learning models, where the first model predicts the viewing direction, and the second model predicts the minimum video prefetch size that can include the actual viewport. We first train two meta models so that they are sensitive to new training data, and then quickly adapt them to users while they are watching the videos. Evaluation results reveal that the meta models can adapt quickly to each user, and can significantly increase the prediction accuracy, especially for the worst-performing predictions.

Node classification for graph-structured data aims to classify nodes whose labels are unknown. While studies on static graphs are prevalent, few studies have focused on dynamic graph node classification. Node classification on dynamic graphs is challenging for two reasons. First, the model needs to capture both structural and temporal information, particularly on dynamic graphs with a long history and require large receptive fields. Second, model scalability becomes a significant concern as the size of the dynamic graph increases. To address these problems, we propose the Time Augmented Dynamic Graph Neural Network (TADGNN) framework. TADGNN consists of two modules: 1) a time augmentation module that captures the temporal evolution of nodes across time structurally, creating a time-augmented spatio-temporal graph, and 2) an information propagation module that learns the dynamic representations for each node across time using the constructed time-augmented graph. We perform node classification experiments on four dynamic graph benchmarks. Experimental results demonstrate that TADGNN framework outperforms several static and dynamic state-of-the-art (SOTA) GNN models while demonstrating superior scalability. We also conduct theoretical and empirical analyses to validate the efficiency of the proposed method. Our code is available at https://sites.google.com/view/tadgnn.

Causal phenomena associated with rare events frequently occur across a wide range of engineering and mathematical problems, such as risk-sensitive safety analysis, accident analysis and prevention, and extreme value theory. However, current methods for causal discovery are often unable to uncover causal links between random variables that manifest only when the variables first experience low-probability realizations. To address this issue, we introduce a novel algorithm that performs statistical independence tests on data collected from time-invariant dynamical systems in which rare but consequential events occur. We seek to understand if the state of the dynamical system causally affects the likelihood of the rare event. In particular, we exploit the time-invariance of the underlying data to superimpose the occurrences of rare events, thus creating a new dataset, with rare events are better represented, on which conditional independence tests can be more efficiently performed. We provide non-asymptotic bounds for the consistency of our algorithm, and validate the performance of our algorithm across various simulated scenarios, with applications to traffic accidents.

面向目标的生成脚本学习旨在根据目标生成后续步骤，这是帮助机器人进行日常生活的刻板印象活动的重要任务。我们表明，如果历史状态不仅被给人的语言指示捕获，而且还可以增强随附图像提供的其他信息，可以提高此任务的性能。因此，我们提出了一项新任务，多媒体生成脚本学习，以通过跟踪文本和视觉方式中的历史状态，并介绍包含2,338个任务和31,496个步骤的第一个基准，从而生成后续步骤。我们旨在生成视觉状态的脚本，这些脚本是可跟踪的，对看不见的任务的诱导性，并且在各自的步骤中多样化。我们建议通过多媒体选择性编码器编码视觉状态更改，并使用检索仪的解码器从先前观察到的任务中转移知识，并通过优化面向多样性的对比度学习目标来在每个步骤中介绍不同的信息。我们定义指标以评估发电质量和电感质量。实验结果表明，我们的方法明显优于强质基线。

强化学习（RL）文献的最新进展使机器人主义者能够在模拟环境中自动训练复杂的政策。但是，由于这些方法的样本复杂性差，使用现实世界数据解决强化学习问题仍然是一个具有挑战性的问题。本文介绍了一种新颖的成本整形方法，旨在减少学习稳定控制器所需的样品数量。该方法添加了一个涉及控制Lyapunov功能（CLF）的术语 - 基于模型的控制文献的“能量样”功能 - 到典型的成本配方。理论结果表明，新的成本会导致使用较小的折现因子时稳定控制器，这是众所周知的，以降低样品复杂性。此外，通过确保即使是高度亚最佳的策略也可以稳定系统，添加CLF术语“鲁棒化”搜索稳定控制器。我们通过两个硬件示例演示了我们的方法，在其中我们学习了一个cartpole的稳定控制器和仅使用几秒钟和几分钟的微调数据的A1稳定控制器。

三维（3D）医学图像的产生可能具有巨大的应用潜力，因为它考虑了3D解剖结构。但是，有两个问题可以防止有效培训3D医疗生成模型：（1）3D医学图像的获取和注释非常昂贵，导致培训图像不足，（2）大量参数是参与3D卷积。为了解决这两个问题，我们提出了一种名为3D Split＆Shuffle-Gan的新型GAN模型。为了解决3D数据稀缺问题，我们首先使用丰富的图像切片预先培训二维（2D）GAN模型，并夸大2D卷积权重以改善3D GAN的初始化。为GAN模型的生成器和鉴别器提出了新型的3D网络体系结构，以显着减少参数的数量，同时保持图像生成的质量。研究了许多体重通胀策略和参数有效的3D架构。对心脏（Stanford Aimi冠状动脉钙）和大脑（阿尔茨海默氏病神经成像计划）的实验表明，所提出的方法会导致改善的3D图像产生质量，参数较少。

已经对机器学习技术进行了广泛的研究，以实现掩盖优化问题，旨在提高掩模的可打印性，更短的周转时间，更好的遮罩制造性等。但是，这些研究中的大多数都集中在小型设计区域的初始解决方案生成上。为了进一步实现机器学习技术在面罩优化任务上的潜力，我们提出了一个卷积傅立叶神经操作员（CFNO），该神经操作员（CFNO）可以有效地学习布局瓷砖依赖性，从而有望使用有限的遗产工具干预，并有望使用无针迹的大规模掩蔽优化。我们在解决非凸优化问题时通过训练有素的机器学习模型发现了岩石引导的自我训练（LGST）的可能性，从而允许迭代模型和数据集更新并带来显着的模型性能改进。实验结果表明，我们基于机器学习的框架首次优于最先进的学术数值掩码优化器，并具有速度级的速度。