Depth estimation from images serves as the fundamental step of 3D perception for autonomous driving and is an economical alternative to expensive depth sensors like LiDAR. The temporal photometric consistency enables self-supervised depth estimation without labels, further facilitating its application. However, most existing methods predict the depth solely based on each monocular image and ignore the correlations among multiple surrounding cameras, which are typically available for modern self-driving vehicles. In this paper, we propose a SurroundDepth method to incorporate the information from multiple surrounding views to predict depth maps across cameras. Specifically, we employ a joint network to process all the surrounding views and propose a cross-view transformer to effectively fuse the information from multiple views. We apply cross-view self-attention to efficiently enable the global interactions between multi-camera feature maps. Different from self-supervised monocular depth estimation, we are able to predict real-world scales given multi-camera extrinsic matrices. To achieve this goal, we adopt structure-from-motion to extract scale-aware pseudo depths to pretrain the models. Further, instead of predicting the ego-motion of each individual camera, we estimate a universal ego-motion of the vehicle and transfer it to each view to achieve multi-view consistency. In experiments, our method achieves the state-of-the-art performance on the challenging multi-camera depth estimation datasets DDAD and nuScenes.
In this paper, we propose the LiDAR Distillation to bridge the domain gap induced by different LiDAR beams for 3D object detection. In many real-world applications, the LiDAR points used by mass-produced robots and vehicles usually have fewer beams than that in large-scale public datasets. Moreover, as the LiDARs are upgraded to other product models with different beam amount, it becomes challenging to utilize the labeled data captured by previous versions' high-resolution sensors. Despite the recent progress on domain adaptive 3D detection, most methods struggle to eliminate the beam-induced domain gap. We find that it is essential to align the point cloud density of the source domain with that of the target domain during the training process. Inspired by this discovery, we propose a progressive framework to mitigate the beam-induced domain shift. In each iteration, we first generate low-beam pseudo LiDAR by downsampling the high-beam point clouds. Then the teacher-student framework is employed to distill rich information from the data with more beams. Extensive experiments on Waymo, nuScenes and KITTI datasets with three different LiDAR-based detectors demonstrate the effectiveness of our LiDAR Distillation. Notably, our approach does not increase any additional computation cost for inference.
For intelligent reflecting surface (IRS)-aided wireless communications, channel estimation is essential and usually requires excessive channel training overhead when the number of IRS reflecting elements is large. The acquisition of accurate channel state information (CSI) becomes more challenging when the channel is not quasi-static due to the mobility of the transmitter and/or receiver. In this work, we study an IRS-aided wireless communication system with a time-varying channel model and propose an innovative two-stage transmission protocol. In the first stage, we send pilot symbols and track the direct/reflected channels based on the received signal, and then data signals are transmitted. In the second stage, instead of sending pilot symbols first, we directly predict the direct/reflected channels and all the time slots are used for data transmission. Based on the proposed transmission protocol, we propose a two-stage channel tracking and prediction (2SCTP) scheme to obtain the direct and reflected channels with low channel training overhead, which is achieved by exploiting the temporal correlation of the time-varying channels. Specifically, we first consider a special case where the IRS-access point (AP) channel is assumed to be static, for which a Kalman filter (KF)-based algorithm and a long short-term memory (LSTM)-based neural network are proposed for channel tracking and prediction, respectively. Then, for the more general case where the IRS-AP, user-IRS and user-AP channels are all assumed to be time-varying, we present a generalized KF (GKF)-based channel tracking algorithm, where proper approximations are employed to handle the underlying non-Gaussian random variables. Numerical simulations are provided to verify the effectiveness of our proposed transmission protocol and channel tracking/prediction algorithms as compared to existing ones.
In this paper, we introduce a deep multi-view stereo (MVS) system that jointly predicts depths, surface normals and per-view confidence maps. The key to our approach is a novel solver that iteratively solves for per-view depth map and normal map by optimizing an energy potential based on the locally planar assumption. Specifically, the algorithm updates depth map by propagating from neighboring pixels with slanted planes, and updates normal map with local probabilistic plane fitting. Both two steps are monitored by a customized confidence map. This solver is not only effective as a post-processing tool for plane-based depth refinement and completion, but also differentiable such that it can be efficiently integrated into deep learning pipelines. Our multi-view stereo system employs multiple optimization steps of the solver over the initial prediction of depths and surface normals. The whole system can be trained end-to-end, decoupling the challenging problem of matching pixels within poorly textured regions from the cost-volume based neural network. Experimental results on ScanNet and RGB-D Scenes V2 demonstrate state-of-the-art performance of the proposed deep MVS system on multi-view depth estimation, with our proposed solver consistently improving the depth quality over both conventional and deep learning based MVS pipelines. Code is available at https://github.com/thuzhaowang/idn-solver.
Cross-Lingual Information Retrieval (CLIR) aims to rank the documents written in a language different from the user's query. The intrinsic gap between different languages is an essential challenge for CLIR. In this paper, we introduce the multilingual knowledge graph (KG) to the CLIR task due to the sufficient information of entities in multiple languages. It is regarded as a "silver bullet" to simultaneously perform explicit alignment between queries and documents and also broaden the representations of queries. And we propose a model named CLIR with hierarchical knowledge enhancement (HIKE) for our task. The proposed model encodes the textual information in queries, documents and the KG with multilingual BERT, and incorporates the KG information in the query-document matching process with a hierarchical information fusion mechanism. Particularly, HIKE first integrates the entities and their neighborhood in KG into query representations with a knowledge-level fusion, then combines the knowledge from both source and target languages to further mitigate the linguistic gap with a language-level fusion. Finally, experimental results demonstrate that HIKE achieves substantial improvements over state-of-the-art competitors.
In this work, we present a new multi-view depth estimation method that utilizes both conventional SfM reconstruction and learning-based priors over the recently proposed neural radiance fields (NeRF). Unlike existing neural network based optimization method that relies on estimated correspondences, our method directly optimizes over implicit volumes, eliminating the challenging step of matching pixels in indoor scenes. The key to our approach is to utilize the learning-based priors to guide the optimization process of NeRF. Our system firstly adapts a monocular depth network over the target scene by finetuning on its sparse SfM reconstruction. Then, we show that the shape-radiance ambiguity of NeRF still exists in indoor environments and propose to address the issue by employing the adapted depth priors to monitor the sampling process of volume rendering. Finally, a per-pixel confidence map acquired by error computation on the rendered image can be used to further improve the depth quality. Experiments show that our proposed framework significantly outperforms state-of-the-art methods on indoor scenes, with surprising findings presented on the effectiveness of correspondence-based optimization and NeRF-based optimization over the adapted depth priors. In addition, we show that the guided optimization scheme does not sacrifice the original synthesis capability of neural radiance fields, improving the rendering quality on both seen and novel views. Code is available at https://github.com/weiyithu/NerfingMVS.
3D point cloud understanding has made great progress in recent years. However, one major bottleneck is the scarcity of annotated real datasets, especially compared to 2D object detection tasks, since a large amount of labor is involved in annotating the real scans of a scene. A promising solution to this problem is to make better use of the synthetic dataset, which consists of CAD object models, to boost the learning on real datasets. This can be achieved by the pre-training and fine-tuning procedure. However, recent work on 3D pre-training exhibits failure when transfer features learned on synthetic objects to other real-world applications. In this work, we put forward a new method called RandomRooms to accomplish this objective. In particular, we propose to generate random layouts of a scene by making use of the objects in the synthetic CAD dataset and learn the 3D scene representation by applying object-level contrastive learning on two random scenes generated from the same set of synthetic objects. The model pre-trained in this way can serve as a better initialization when later fine-tuning on the 3D object detection task. Empirically, we show consistent improvement in downstream 3D detection tasks on several base models, especially when less training data are used, which strongly demonstrates the effectiveness and generalization of our method. Benefiting from the rich semantic knowledge and diverse objects from synthetic data, our method establishes the new state-of-the-art on widely-used 3D detection benchmarks ScanNetV2 and SUN RGB-D. We expect our attempt to provide a new perspective for bridging object and scene-level 3D understanding.
In this paper, we investigate the problem of weakly supervised 3D vehicle detection. Conventional methods for 3D object detection need vast amounts of manually labelled 3D data as supervision signals. However, annotating large datasets requires huge human efforts, especially for 3D area. To tackle this problem, we propose frustum-aware geometric reasoning (FGR) to detect vehicles in point clouds without any 3D annotations. Our method consists of two stages: coarse 3D segmentation and 3D bounding box estimation. For the first stage, a context-aware adaptive region growing algorithm is designed to segment objects based on 2D bounding boxes. Leveraging predicted segmentation masks, we develop an anti-noise approach to estimate 3D bounding boxes in the second stage. Finally 3D pseudo labels generated by our method are utilized to train a 3D detector. Independent of any 3D groundtruth, FGR reaches comparable performance with fully supervised methods on the KITTI dataset. The findings indicate that it is able to accurately detect objects in 3D space with only 2D bounding boxes and sparse point clouds.
Training a deep neural network heavily relies on a large amount of training data with accurate annotations. To alleviate this problem, various methods have been proposed to annotate the data automatically. However, automatically generating annotations will inevitably yields noisy labels. In this paper, we propose a Data Selection and joint Training (DST) method to automatically select training samples with accurate annotations. Specifically, DST fits a mixture model according to the original annotation as well as the predicted label for each training sample, and the mixture model is utilized to dynamically divide the training dataset into a correctly labeled dataset, a correctly predicted set and a wrong dataset. Then, DST is trained with these datasets in a supervised manner. Due to confirmation bias problem, we train the two networks alternately, and each network is tasked to establish the data division to teach another network. For each iteration, the correctly labeled and predicted labels are reweighted respectively by the probabilities from the mixture model, and a uniform distribution is used to generate the probabilities of the wrong samples. Experiments on CIFAR-10, CIFAR-100 and Clothing1M demonstrate that DST is the comparable or superior to the state-of-the-art methods.
A bipartite network is a graph structure where nodes are from two distinct domains and only inter-domain interactions exist as edges. A large number of network embedding methods exist to learn vectorial node representations from general graphs with both homogeneous and heterogeneous node and edge types, including some that can specifically model the distinct properties of bipartite networks. However, these methods are inadequate to model multiplex bipartite networks (e.g., in e-commerce), that have multiple types of interactions (e.g., click, inquiry, and buy) and node attributes. Most real-world multiplex bipartite networks are also sparse and have imbalanced node distributions that are challenging to model. In this paper, we develop an unsupervised Dual HyperGraph Convolutional Network (DualHGCN) model that scalably transforms the multiplex bipartite network into two sets of homogeneous hypergraphs and uses spectral hypergraph convolutional operators, along with intra- and inter-message passing strategies to promote information exchange within and across domains, to learn effective node embedding. We benchmark DualHGCN using four real-world datasets on link prediction and node classification tasks. Our extensive experiments demonstrate that DualHGCN significantly outperforms state-of-the-art methods, and is robust to varying sparsity levels and imbalanced node distributions.