Yi Yang

Yu Gao, Lutong Su, Hao Liang et al.

Neural Radiance Fields (NeRF) use multi-view images for 3D scene representation, demonstrating remarkable performance. As one of the primary sources of multi-view images, multi-camera systems encounter challenges such as varying intrinsic parameters and frequent pose changes. Most previous NeRF-based methods assume a unique camera and rarely consider multi-camera scenarios. Besides, some NeRF methods that can optimize intrinsic and extrinsic parameters still remain susceptible to suboptimal solutions when these parameters are poor initialized. In this paper, we propose MC-NeRF, a method that enables joint optimization of both intrinsic and extrinsic parameters alongside NeRF. The method also supports each image corresponding to independent camera parameters. First, we tackle coupling issue and the degenerate case that arise from the joint optimization between intrinsic and extrinsic parameters. Second, based on the proposed solutions, we introduce an efficient calibration image acquisition scheme for multi-camera systems, including the design of calibration object. Finally, we present an end-to-end network with training sequence that enables the estimation of intrinsic and extrinsic parameters, along with the rendering network. Furthermore, recognizing that most existing datasets are designed for a unique camera, we construct a real multi-camera image acquisition system and create a corresponding new dataset, which includes both simulated data and real-world captured images. Experiments confirm the effectiveness of our method when each image corresponds to different camera parameters. Specifically, we use multi-cameras, each with different intrinsic and extrinsic parameters in real-world system, to achieve 3D scene representation without providing initial poses.

Huiyao Chen, Yi Yang, Yinghui Li et al.

Existing long-document question answering systems typically process texts as flat sequences or use heuristic chunking, which overlook the discourse structures that naturally guide human comprehension. We present a discourse-aware hierarchical framework that leverages rhetorical structure theory (RST) for long document question answering. Our approach converts discourse trees into sentence-level representations and employs LLM-enhanced node representations to bridge structural and semantic information. The framework involves three key innovations: language-universal discourse parsing for lengthy documents, LLM-based enhancement of discourse relation nodes, and structure-guided hierarchical retrieval. Extensive experiments on four datasets demonstrate consistent improvements over existing approaches through the incorporation of discourse structure, across multiple genres and languages. Moreover, the proposed framework exhibits strong robustness across diverse document types and linguistic settings.

Yang Wang, Zhenheng Yang, Peng Wang et al.

Learning depth and optical flow via deep neural networks by watching videos has made significant progress recently. In this paper, we jointly solve the two tasks by exploiting the underlying geometric rules within stereo videos. Specifically, given two consecutive stereo image pairs from a video, we first estimate depth, camera ego-motion and optical flow from three neural networks. Then the whole scene is decomposed into moving foreground and static background by compar- ing the estimated optical flow and rigid flow derived from the depth and ego-motion. We propose a novel consistency loss to let the optical flow learn from the more accurate rigid flow in static regions. We also design a rigid alignment module which helps refine ego-motion estimation by using the estimated depth and optical flow. Experiments on the KITTI dataset show that our results significantly outperform other state-of- the-art algorithms. Source codes can be found at https: //github.com/baidu-research/UnDepthflow

Shuchang Zhou, Taihong Xiao, Yi Yang et al.

Object Transfiguration replaces an object in an image with another object from a second image. For example it can perform tasks like "putting exactly those eyeglasses from image A on the nose of the person in image B". Usage of exemplar images allows more precise specification of desired modifications and improves the diversity of conditional image generation. However, previous methods that rely on feature space operations, require paired data and/or appearance models for training or disentangling objects from background. In this work, we propose a model that can learn object transfiguration from two unpaired sets of images: one set containing images that "have" that kind of object, and the other set being the opposite, with the mild constraint that the objects be located approximately at the same place. For example, the training data can be one set of reference face images that have eyeglasses, and another set of images that have not, both of which spatially aligned by face landmarks. Despite the weak 0/1 labels, our model can learn an "eyeglasses" subspace that contain multiple representatives of different types of glasses. Consequently, we can perform fine-grained control of generated images, like swapping the glasses in two images by swapping the projected components in the "eyeglasses" subspace, to create novel images of people wearing eyeglasses. Overall, our deterministic generative model learns disentangled attribute subspaces from weakly labeled data by adversarial training. Experiments on CelebA and Multi-PIE datasets validate the effectiveness of the proposed model on real world data, in generating images with specified eyeglasses, smiling, hair styles, and lighting conditions etc. The code is available online.

Zhiyuan Min, Yawei Luo, Jianwen Sun et al.

Generalizable 3D Gaussian splitting (3DGS) can reconstruct new scenes from sparse-view observations in a feed-forward inference manner, eliminating the need for scene-specific retraining required in conventional 3DGS. However, existing methods rely heavily on epipolar priors, which can be unreliable in complex realworld scenes, particularly in non-overlapping and occluded regions. In this paper, we propose eFreeSplat, an efficient feed-forward 3DGS-based model for generalizable novel view synthesis that operates independently of epipolar line constraints. To enhance multiview feature extraction with 3D perception, we employ a selfsupervised Vision Transformer (ViT) with cross-view completion pre-training on large-scale datasets. Additionally, we introduce an Iterative Cross-view Gaussians Alignment method to ensure consistent depth scales across different views. Our eFreeSplat represents an innovative approach for generalizable novel view synthesis. Different from the existing pure geometry-free methods, eFreeSplat focuses more on achieving epipolar-free feature matching and encoding by providing 3D priors through cross-view pretraining. We evaluate eFreeSplat on wide-baseline novel view synthesis tasks using the RealEstate10K and ACID datasets. Extensive experiments demonstrate that eFreeSplat surpasses state-of-the-art baselines that rely on epipolar priors, achieving superior geometry reconstruction and novel view synthesis quality. Project page: https://tatakai1.github.io/efreesplat/.

Xihang Yue, Yi Yang, Linchao Zhu

Recent advances in operator learning have produced two distinct approaches for solving partial differential equations (PDEs): attention-based methods offering point-level adaptability but lacking spectral constraints, and spectral-based methods providing domain-level continuity priors but limited in local flexibility. This dichotomy has hindered the development of PDE solvers with both strong flexibility and generalization capability. This work introduces Holistic Physics Mixer (HPM), a simple framework that bridges this gap by integrating spectral and physical information in a unified space. HPM unifies both approaches as special cases while enabling more powerful spectral-physical interactions beyond either method alone. This enables HPM to inherit both the strong generalization of spectral methods and the flexibility of attention mechanisms while avoiding their respective limitations. Through extensive experiments across diverse PDE problems, we demonstrate that HPM consistently outperforms state-of-the-art methods in both accuracy and computational efficiency, while maintaining strong generalization capabilities with limited training data and excellent zero-shot performance on unseen resolutions.

Hao Liang, Zhipeng Dong, Kaixin Chen et al.

Surround-view perception has garnered significant attention for its ability to enhance the perception capabilities of autonomous driving vehicles through the exchange of information with surrounding cameras. However, existing surround-view perception systems are limited by inefficiencies in unidirectional interaction pattern with human and distortions in overlapping regions exponentially propagating into non-overlapping areas. To address these challenges, this paper introduces ChatStitch, a surround-view human-machine co-perception system capable of unveiling obscured blind spot information through natural language commands integrated with external digital assets. To dismantle the unidirectional interaction bottleneck, ChatStitch implements a cognitively grounded closed-loop interaction multi-agent framework based on Large Language Models. To suppress distortion propagation across overlapping boundaries, ChatStitch proposes SV-UDIS, a surround-view unsupervised deep image stitching method under the non-global-overlapping condition. We conducted extensive experiments on the UDIS-D, MCOV-SLAM open datasets, and our real-world dataset. Specifically, our SV-UDIS method achieves state-of-the-art performance on the UDIS-D dataset for 3, 4, and 5 image stitching tasks, with PSNR improvements of 9\%, 17\%, and 21\%, and SSIM improvements of 8\%, 18\%, and 26\%, respectively.

Yaxiong Wang, Yunchao Wei, Xueming Qian et al.

We aim to tackle the challenging yet practical scenery image outpainting task in this work. Recently, generative adversarial learning has significantly advanced the image outpainting by producing semantic consistent content for the given image. However, the existing methods always suffer from the blurry texture and the artifacts of the generative part, making the overall outpainting results lack authenticity. To overcome the weakness, this work investigates a principle way to synthesize texture-rich results by borrowing pixels from its neighbors (i.e., reference images), named \textbf{Re}ference-\textbf{G}uided \textbf{O}utpainting (ReGO). Particularly, the ReGO designs an Adaptive Content Selection (ACS) module to transfer the pixel of reference images for texture compensating of the target one. To prevent the style of the generated part from being affected by the reference images, a style ranking loss is further proposed to augment the ReGO to synthesize style-consistent results. Extensive experiments on two popular benchmarks, NS6K \cite{yangzx} and NS8K \cite{wang}, well demonstrate the effectiveness of our ReGO. Our code will be made public available.

Tong Liu, Zhaowei Chen, Yi Yang et al.

Nowadays, deep learning techniques are widely used for lane detection, but application in low-light conditions remains a challenge until this day. Although multi-task learning and contextual-information-based methods have been proposed to solve the problem, they either require additional manual annotations or introduce extra inference overhead respectively. In this paper, we propose a style-transfer-based data enhancement method, which uses Generative Adversarial Networks (GANs) to generate images in low-light conditions, that increases the environmental adaptability of the lane detector. Our solution consists of three parts: the proposed SIM-CycleGAN, light conditions style transfer and lane detection network. It does not require additional manual annotations nor extra inference overhead. We validated our methods on the lane detection benchmark CULane using ERFNet. Empirically, lane detection model trained using our method demonstrated adaptability in low-light conditions and robustness in complex scenarios. Our code for this paper will be publicly available.

Yi Yang, Baile Xu, Furao Shen et al.

User response prediction makes a crucial contribution to the rapid development of online advertising system and recommendation system. The importance of learning feature interactions has been emphasized by many works. Many deep models are proposed to automatically learn high-order feature interactions. Since most features in advertising system and recommendation system are high-dimensional sparse features, deep models usually learn a low-dimensional distributed representation for each feature in the bottom layer. Besides traditional fully-connected architectures, some new operations, such as convolutional operations and product operations, are proposed to learn feature interactions better. In these models, the representation is shared among different operations. However, the best representation for different operations may be different. In this paper, we propose a new neural model named Operation-aware Neural Networks (ONN) which learns different representations for different operations. Our experimental results on two large-scale real-world ad click/conversion datasets demonstrate that ONN consistently outperforms the state-of-the-art models in both offline-training environment and online-training environment.