Image super-resolution (SR) is a technique to recover lost high-frequency information in low-resolution (LR) images. Spatial-domain information has been widely exploited to implement image SR, so a new trend is to involve frequency-domain information in SR tasks. Besides, image SR is typically application-oriented and various computer vision tasks call for image arbitrary magnification. Therefore, in this paper, we study image features in the frequency domain to design a novel scale-arbitrary image SR network. First, we statistically analyze LR-HR image pairs of several datasets under different scale factors and find that the high-frequency spectra of different images under different scale factors suffer from different degrees of degradation, but the valid low-frequency spectra tend to be retained within a certain distribution range. Then, based on this finding, we devise an adaptive scale-aware feature division mechanism using deep reinforcement learning, which can accurately and adaptively divide the frequency spectrum into the low-frequency part to be retained and the high-frequency one to be recovered. Finally, we design a scale-aware feature recovery module to capture and fuse multi-level features for reconstructing the high-frequency spectrum at arbitrary scale factors. Extensive experiments on public datasets show the superiority of our method compared with state-of-the-art methods.

Multivariate time series forecasting with hierarchical structure is pervasive in real-world applications, demanding not only predicting each level of the hierarchy, but also reconciling all forecasts to ensure coherency, i.e., the forecasts should satisfy the hierarchical aggregation constraints. Moreover, the disparities of statistical characteristics between levels can be huge, worsened by non-Gaussian distributions and non-linear correlations. To this extent, we propose a novel end-to-end hierarchical time series forecasting model, based on conditioned normalizing flow-based autoregressive transformer reconciliation, to represent complex data distribution while simultaneously reconciling the forecasts to ensure coherency. Unlike other state-of-the-art methods, we achieve the forecasting and reconciliation simultaneously without requiring any explicit post-processing step. In addition, by harnessing the power of deep model, we do not rely on any assumption such as unbiased estimates or Gaussian distribution. Our evaluation experiments are conducted on four real-world hierarchical datasets from different industrial domains (three public ones and a dataset from the application servers of Alipay's data center) and the preliminary results demonstrate efficacy of our proposed method.

近年来，隐性神经表示（INR）已经出现并显示了其比离散表示的好处。但是，将INR拟合到给定的观察结果通常需要从头开始优化，这是效率低下的，并且在稀疏观测值中概括不佳。为了解决这个问题，大多数先前的作品都会训练一个生成单个向量来调节INR权重的超网络，其中单个向量成为限制输出INR重建精度的信息瓶颈。最近的工作表明，可以通过基于梯度的元学习来精确地推断出INR中的整个权重。由基于梯度的元学习的广义公式的动机，我们提出了一种使用变压器作为INRS的超网络的公式，它可以直接使用专门为设置映射的变压器构建整个INR权重。我们证明了我们在不同任务和域中构建INR的方法的有效性，包括2D图像回归和3D对象的查看合成。我们的工作吸引了变压器超网与基于梯度的元学习算法之间的连接，我们提供了进一步的分析，以理解生成的INRS。带代码的项目页面位于\ url {https://yinboc.github.io/trans-inr/}。

在语义细分中，将高级上下文信息与低级详细信息集成至关重要。为此，大多数现有的分割模型都采用双线性启动采样和卷积来具有不同尺度的地图，然后以相同的分辨率对齐。但是，双线性启动采样模糊了这些特征地图和卷积中所学到的精确信息，这会产生额外的计算成本。为了解决这些问题，我们提出了隐式特征对齐函数（IFA）。我们的方法的灵感来自隐式神经表示的快速扩展的主题，在该主题中，基于坐标的神经网络用于指定信号字段。在IFA中，特征向量被视为表示2D信息字段。给定查询坐标，附近的具有相对坐标的特征向量是从多级特征图中获取的，然后馈入MLP以生成相应的输出。因此，IFA隐含地将特征图在不同级别对齐，并能够在任意分辨率中产生分割图。我们证明了IFA在多个数据集上的功效，包括CityScapes，Pascal环境和ADE20K。我们的方法可以与各种体系结构的改进结合使用，并在共同基准上实现最新的计算准确性权衡。代码将在https://github.com/hzhupku/ifa上提供。

视频通常将流和连续的视觉数据记录为离散的连续帧。由于存储成本对于高保真度的视频来说是昂贵的，因此大多数存储以相对较低的分辨率和帧速率存储。最新的时空视频超分辨率（STVSR）的工作是开发出来的，以将时间插值和空间超分辨率纳入统一框架。但是，其中大多数仅支持固定的上采样量表，这限制了其灵活性和应用。在这项工作中，我们没有遵循离散表示，我们提出了视频隐式神经表示（videoinr），并显示了其对STVSR的应用。学到的隐式神经表示可以解码为任意空间分辨率和帧速率的视频。我们表明，Videoinr在常见的上采样量表上使用最先进的STVSR方法实现了竞争性能，并且在连续和训练的分布量表上显着优于先前的作品。我们的项目页面位于http://zeyuan-chen.com/videoinr/。

How to represent an image? While the visual world is presented in a continuous manner, machines store and see the images in a discrete way with 2D arrays of pixels. In this paper, we seek to learn a continuous representation for images. Inspired by the recent progress in 3D reconstruction with implicit neural representation, we propose Local Implicit Image Function (LIIF), which takes an image coordinate and the 2D deep features around the coordinate as inputs, predicts the RGB value at a given coordinate as an output. Since the coordinates are continuous, LIIF can be presented in arbitrary resolution. To generate the continuous representation for images, we train an encoder with LIIF representation via a self-supervised task with superresolution. The learned continuous representation can be presented in arbitrary resolution even extrapolate to ×30 higher resolution, where the training tasks are not provided. We further show that LIIF representation builds a bridge between discrete and continuous representation in 2D, it naturally supports the learning tasks with size-varied image ground-truths and significantly outperforms the method with resizing the ground-truths. Our project page with code is at https://yinboc.github.io/liif/.

Unsupervised domain adaptation (UDA) for semantic segmentation is a promising task freeing people from heavy annotation work. However, domain discrepancies in low-level image statistics and high-level contexts compromise the segmentation performance over the target domain. A key idea to tackle this problem is to perform both image-level and feature-level adaptation jointly. Unfortunately, there is a lack of such unified approaches for UDA tasks in the existing literature. This paper proposes a novel UDA pipeline for semantic segmentation that unifies image-level and feature-level adaptation. Concretely, for image-level domain shifts, we propose a global photometric alignment module and a global texture alignment module that align images in the source and target domains in terms of image-level properties. For feature-level domain shifts, we perform global manifold alignment by projecting pixel features from both domains onto the feature manifold of the source domain; and we further regularize category centers in the source domain through a category-oriented triplet loss and perform target domain consistency regularization over augmented target domain images. Experimental results demonstrate that our pipeline significantly outperforms previous methods. In the commonly tested GTA5$\rightarrow$Cityscapes task, our proposed method using Deeplab V3+ as the backbone surpasses previous SOTA by 8%, achieving 58.2% in mIoU.

Different people speak with diverse personalized speaking styles. Although existing one-shot talking head methods have made significant progress in lip sync, natural facial expressions, and stable head motions, they still cannot generate diverse speaking styles in the final talking head videos. To tackle this problem, we propose a one-shot style-controllable talking face generation framework. In a nutshell, we aim to attain a speaking style from an arbitrary reference speaking video and then drive the one-shot portrait to speak with the reference speaking style and another piece of audio. Specifically, we first develop a style encoder to extract dynamic facial motion patterns of a style reference video and then encode them into a style code. Afterward, we introduce a style-controllable decoder to synthesize stylized facial animations from the speech content and style code. In order to integrate the reference speaking style into generated videos, we design a style-aware adaptive transformer, which enables the encoded style code to adjust the weights of the feed-forward layers accordingly. Thanks to the style-aware adaptation mechanism, the reference speaking style can be better embedded into synthesized videos during decoding. Extensive experiments demonstrate that our method is capable of generating talking head videos with diverse speaking styles from only one portrait image and an audio clip while achieving authentic visual effects. Project Page: https://github.com/FuxiVirtualHuman/styletalk.

Witnessing the impressive achievements of pre-training techniques on large-scale data in the field of computer vision and natural language processing, we wonder whether this idea could be adapted in a grab-and-go spirit, and mitigate the sample inefficiency problem for visuomotor driving. Given the highly dynamic and variant nature of the input, the visuomotor driving task inherently lacks view and translation invariance, and the visual input contains massive irrelevant information for decision making, resulting in predominant pre-training approaches from general vision less suitable for the autonomous driving task. To this end, we propose PPGeo (Policy Pre-training via Geometric modeling), an intuitive and straightforward fully self-supervised framework curated for the policy pretraining in visuomotor driving. We aim at learning policy representations as a powerful abstraction by modeling 3D geometric scenes on large-scale unlabeled and uncalibrated YouTube driving videos. The proposed PPGeo is performed in two stages to support effective self-supervised training. In the first stage, the geometric modeling framework generates pose and depth predictions simultaneously, with two consecutive frames as input. In the second stage, the visual encoder learns driving policy representation by predicting the future ego-motion and optimizing with the photometric error based on current visual observation only. As such, the pre-trained visual encoder is equipped with rich driving policy related representations and thereby competent for multiple visuomotor driving tasks. Extensive experiments covering a wide span of challenging scenarios have demonstrated the superiority of our proposed approach, where improvements range from 2% to even over 100% with very limited data. Code and models will be available at https://github.com/OpenDriveLab/PPGeo.

Increasing research interests focus on sequential recommender systems, aiming to model dynamic sequence representation precisely. However, the most commonly used loss function in state-of-the-art sequential recommendation models has essential limitations. To name a few, Bayesian Personalized Ranking (BPR) loss suffers the vanishing gradient problem from numerous negative sampling and predictionbiases; Binary Cross-Entropy (BCE) loss subjects to negative sampling numbers, thereby it is likely to ignore valuable negative examples and reduce the training efficiency; Cross-Entropy (CE) loss only focuses on the last timestamp of the training sequence, which causes low utilization of sequence information and results in inferior user sequence representation. To avoid these limitations, in this paper, we propose to calculate Cumulative Cross-Entropy (CCE) loss over the sequence. CCE is simple and direct, which enjoys the virtues of painless deployment, no negative sampling, and effective and efficient training. We conduct extensive experiments on five benchmark datasets to demonstrate the effectiveness and efficiency of CCE. The results show that employing CCE loss on three state-of-the-art models GRU4Rec, SASRec, and S3-Rec can reach 125.63%, 69.90%, and 33.24% average improvement of full ranking NDCG@5, respectively. Using CCE, the performance curve of the models on the test data increases rapidly with the wall clock time, and is superior to that of other loss functions in almost the whole process of model training.