Stereo video synthesis from monocular input is challenging in spatial computing and virtual reality due to the lack of high-quality stereo video pairs for training and the difficulty of maintaining spatio-temporal consistency between frames. Existing methods primarily address these issues by directly applying novel view synthesis (NVS) techniques to video, while facing limitations such as the inability to effectively represent dynamic scenes and the requirement for extensive training data. In this paper, we introduce a novel self-supervised stereo video synthesis paradigm via a video diffusion model, termed SpatialDreamer, which meets the challenges head-on. Firstly, to address the stereo video data insufficiency, we propose a Depth based Video Generation module DVG, which employs a forward-backward rendering mechanism to generate paired videos with geometric and temporal priors. Leveraging data generated by DVG, we propose RefinerNet along with a self-supervised synthetic framework designed to facilitate efficient and dedicated training. More importantly, we devise a consistency control module, which consists of a metric of stereo deviation strength and a Temporal Interaction Learning module TIL for geometric and temporal consistency ensurance respectively. We evaluated the proposed method against various benchmark methods, with the results showcasing its superior performance. Our project website is at: https://spatialdreamer.github.io.

Clothing changes and lack of data labels are both crucial challenges in person ReID. For the former challenge, people may occur multiple times wearing different clothes. A majority of current research works for person ReID focuses on benchmarks in which a person’s clothing is kept the same all the time. For the latter challenge, most researchers try to transfer information from a labeled source dataset to an unlabeled target dataset. Whereas purely unsupervised training is less used. In this letter, we aim to solve both problems simultaneously. We design a novel unsupervised model, Syn-Person-Cluster ReID, to solve the unlabeled clothing change person ReID problem. We develop a purely unsupervised pipeline equipped with two innovations: synthetic augmentation on person images and feature restriction for the same person. The synthetic augmentation is to supply additional information for the same person, which can then be used as partially supervised inputs to the feature restriction. Extensive experiments on clothing change ReID datasets show the out-performance of our methods.

Person Re-identification (ReID) is a critical computer vision task which aims to match the same person in images or video sequences. Most current works focus on settings where the resolution of images is kept the same. However, the resolution is a crucial factor in person ReID, especially when the cameras are at different distances from the person or the camera’s models are different from each other. In this paper, we propose a novel two-stream network with a lightweight resolution association ReID feature transformation (RAFT) module and a self-weighted attention (SWA) ReID module to evaluate features under different resolutions. RAFT transforms the low resolution features to corresponding high resolution features. SWA evaluates both features to get weight factors for the person ReID. Both modules are jointly trained to get a resolution-invariant representation. Extensive experiments on five benchmark datasets show the effectiveness of our method. For instance, we achieve Rank-1 accuracy of 43.3% and 83.2% on CAVIAR and MLR-CUHK03, outperforming the state-of-the-art.

The neural network based approach for 3D human pose estimation from monocular images has attracted growing interest. However, annotating 3D poses is a labor-intensive and expensive process. In this paper, we propose a novel self-supervised approach to avoid the need of manual annotations. Different from existing weakly/self-supervised methods that require extra unpaired 3D ground-truth data to alleviate the depth ambiguity problem, our method trains the network only relying on geometric knowledge without any additional 3D pose annotations. The proposed method follows the two-stage pipeline: 2D pose estimation and 2D-to-3D pose lifting. We design the transform re-projection loss that is an effective way to explore multi-view consistency for training the 2D-to-3D lifting network. Besides, we adopt the confidences of 2D joints to integrate losses from different views to alleviate the influence of noises caused by the self-occlusion problem. Finally, we design a two-branch training architecture, which helps to preserve the scale information of re-projected 2D poses during training, resulting in accurate 3D pose predictions. We demonstrate the effectiveness of our method on two popular 3D human pose datasets, Human3.6M and MPI-INF-3DHP. The results show that our method significantly outperforms recent weakly/self-supervised approaches.

Most existing works in Person Re-identification (ReID) focus on settings where illumination either is kept the same or has very little fluctuation. However, the changes in the illumination degree may affect the robustness of a ReID algorithm significantly. To address this problem, we proposed a Two-Stream Network that can separate ReID features from lighting features to enhance ReID performance. Its innovations are threefold: (1) A discriminative entropy loss to ensure the ReID features contain no lighting information. (2) A ReID Teacher model trained by images under “neutral” lighting conditions to guide ReID classification. (3) An illumination Teacher model trained by the differences between the illumination-adjusted and original images to guide illumination classification. We construct two augmented datasets by synthetically changing a set of predefined lighting conditions in two of the most popular ReID benchmarks: Market1501 and DukeMTMC-ReID. Experiments demonstrate that our algorithm outperforms other state-of-the-art works and particularly potent in handling images under extremely low light.