Multiview detection uses multiple calibrated cameras with overlapping fields of views to locate occluded pedestrians. In this field, existing methods typically adopt a ``human modeling - aggregation'' strategy. To find robust pedestrian representations, some intuitively incorporate 2D perception results from each frame, while others use entire frame features projected to the ground plane. However, the former does not consider the human appearance and leads to many ambiguities, and the latter suffers from projection errors due to the lack of accurate height of the human torso and head. In this paper, we propose a new pedestrian representation scheme based on human point clouds modeling. Specifically, using ray tracing for holistic human depth estimation, we model pedestrians as upright, thin cardboard point clouds on the ground. Then, we aggregate the point clouds of the pedestrian cardboard across multiple views for a final decision. Compared with existing representations, the proposed method explicitly leverages human appearance and reduces projection errors significantly by relatively accurate height estimation. On four standard evaluation benchmarks, the proposed method achieves very competitive results. Our code and data will be released at https://github.com/ZichengDuan/MvCHM.
Besides image classification, Contrastive Language-Image Pre-training (CLIP) has accomplished extraordinary success for a wide range of vision tasks, including object-level and 3D space understanding. However, it's still challenging to transfer semantic knowledge learned from CLIP into more intricate tasks of quantified targets, such as depth estimation with geometric information. In this paper, we propose to apply CLIP for zero-shot monocular depth estimation, named DepthCLIP. We found that the patches of the input image could respond to a certain semantic distance token and then be projected to a quantified depth bin for coarse estimation. Without any training, our DepthCLIP surpasses existing unsupervised methods and even approaches the early fully-supervised networks. To our best knowledge, we are the first to conduct zero-shot adaptation from the semantic language knowledge to quantified downstream tasks and perform zero-shot monocular depth estimation. We hope our work could cast a light on future research. The code is available at https://github.com/Adonis-galaxy/DepthCLIP.
In this paper we present a novel self-supervised method to anticipate the depth estimate for a future, unobserved real-world urban scene. This work is the first to explore self-supervised learning for estimation of monocular depth of future unobserved frames of a video. Existing works rely on a large number of annotated samples to generate the probabilistic prediction of depth for unseen frames. However, this makes it unrealistic due to its requirement for large amount of annotated depth samples of video. In addition, the probabilistic nature of the case, where one past can have multiple future outcomes often leads to incorrect depth estimates. Unlike previous methods, we model the depth estimation of the unobserved frame as a view-synthesis problem, which treats the depth estimate of the unseen video frame as an auxiliary task while synthesizing back the views using learned pose. This approach is not only cost effective - we do not use any ground truth depth for training (hence practical) but also deterministic (a sequence of past frames map to an immediate future). To address this task we first develop a novel depth forecasting network DeFNet which estimates depth of unobserved future by forecasting latent features. Second, we develop a channel-attention based pose estimation network that estimates the pose of the unobserved frame. Using this learned pose, estimated depth map is reconstructed back into the image domain, thus forming a self-supervised solution. Our proposed approach shows significant improvements in Abs Rel metric compared to state-of-the-art alternatives on both short and mid-term forecasting setting, benchmarked on KITTI and Cityscapes. Code is available at https://github.com/sauradip/depthForecasting
We introduce MGNet, a multi-task framework for monocular geometric scene understanding. We define monocular geometric scene understanding as the combination of two known tasks: Panoptic segmentation and self-supervised monocular depth estimation. Panoptic segmentation captures the full scene not only semantically, but also on an instance basis. Self-supervised monocular depth estimation uses geometric constraints derived from the camera measurement model in order to measure depth from monocular video sequences only. To the best of our knowledge, we are the first to propose the combination of these two tasks in one single model. Our model is designed with focus on low latency to provide fast inference in real-time on a single consumer-grade GPU. During deployment, our model produces dense 3D point clouds with instance aware semantic labels from single high-resolution camera images. We evaluate our model on two popular autonomous driving benchmarks, i.e., Cityscapes and KITTI, and show competitive performance among other real-time capable methods. Source code is available at https://github.com/markusschoen/MGNet.
Recent works in autonomous driving have widely adopted the bird's-eye-view (BEV) semantic map as an intermediate representation of the world. Online prediction of these BEV maps involves non-trivial operations such as multi-camera data extraction as well as fusion and projection into a common topview grid. This is usually done with error-prone geometric operations (e.g., homography or back-projection from monocular depth estimation) or expensive direct dense mapping between image pixels and pixels in BEV (e.g., with MLP or attention). In this work, we present 'LaRa', an efficient encoder-decoder, transformer-based model for vehicle semantic segmentation from multiple cameras. Our approach uses a system of cross-attention to aggregate information over multiple sensors into a compact, yet rich, collection of latent representations. These latent representations, after being processed by a series of self-attention blocks, are then reprojected with a second cross-attention in the BEV space. We demonstrate that our model outperforms the best previous works using transformers on nuScenes. The code and trained models are available at https://github.com/valeoai/LaRa
Neural networks have shown great success in extracting geometric information from color images. Especially, monocular depth estimation networks are increasingly reliable in real-world scenes. In this work we investigate the applicability of such monocular depth estimation networks to semi-transparent volume rendered images. As depth is notoriously difficult to define in a volumetric scene without clearly defined surfaces, we consider different depth computations that have emerged in practice, and compare state-of-the-art monocular depth estimation approaches for these different interpretations during an evaluation considering different degrees of opacity in the renderings. Additionally, we investigate how these networks can be extended to further obtain color and opacity information, in order to create a layered representation of the scene based on a single color image. This layered representation consists of spatially separated semi-transparent intervals that composite to the original input rendering. In our experiments we show that existing approaches to monocular depth estimation can be adapted to perform well on semi-transparent volume renderings, which has several applications in the area of scientific visualization, like re-composition with additional objects and labels or additional shading.
Accurate depth estimation with lowest compute and energy cost is a crucial requirement for unmanned and battery operated autonomous systems. Robotic applications require real time depth estimation for navigation and decision making under rapidly changing 3D surroundings. A high accuracy algorithm may provide the best depth estimation but may consume tremendous compute and energy resources. A general trade-off is to choose less accurate methods for initial depth estimate and a more accurate yet compute intensive method when needed. Previous work has shown this trade-off can be improved by developing a state-of-the-art method (AnyNet) to improve stereo depth estimation. We studied both the monocular and stereo vision depth estimation methods and investigated methods to reduce computational complexity of these methods. This was our baseline. Consequently, our experiments show reduction of monocular depth estimation model size by ~75% reduces accuracy by less than 2% (SSIM metric). Our experiments with the novel stereo vision method (AnyNet) show that accuracy of depth estimation does not degrade more than 3% (three pixel error metric) in spite of reduction in model size by ~20%. We have shown that smaller models can indeed perform competitively.
In this research, we propose a new 3D object detector with a trustworthy depth estimation, dubbed BEVDepth, for camera-based Bird's-Eye-View (BEV) 3D object detection. Our work is based on a key observation -- depth estimation in recent approaches is surprisingly inadequate given the fact that depth is essential to camera 3D detection. Our BEVDepth resolves this by leveraging explicit depth supervision. A camera-awareness depth estimation module is also introduced to facilitate the depth predicting capability. Besides, we design a novel Depth Refinement Module to counter the side effects carried by imprecise feature unprojection. Aided by customized Efficient Voxel Pooling and multi-frame mechanism, BEVDepth achieves the new state-of-the-art 60.9% NDS on the challenging nuScenes test set while maintaining high efficiency. For the first time, the NDS score of a camera model reaches 60%.
Multi-task learning has recently emerged as a promising solution for a comprehensive understanding of complex scenes. In addition to being memory-efficient, multi-task models, when appropriately designed, can facilitate the exchange of complementary signals across tasks. In this work, we jointly address 2D semantic segmentation and three geometry-related tasks: dense depth estimation, surface normal estimation, and edge estimation, demonstrating their benefits on both indoor and outdoor datasets. We propose a novel multi-task learning architecture that leverages pairwise cross-task exchange through correlation-guided attention and self-attention to enhance the overall representation learning for all tasks. We conduct extensive experiments across three multi-task setups, showing the advantages of our approach compared to competitive baselines in both synthetic and real-world benchmarks. Additionally, we extend our method to the novel multi-task unsupervised domain adaptation setting. Our code is available at https://github.com/cv-rits/DenseMTL
3D hand pose estimation methods have made significant progress recently. However, the estimation accuracy is often far from sufficient for specific real-world applications, and thus there is significant room for improvement. This paper proposes TriHorn-Net, a novel model that uses specific innovations to improve hand pose estimation accuracy on depth images. The first innovation is the decomposition of the 3D hand pose estimation into the estimation of 2D joint locations in the depth image space (UV), and the estimation of their corresponding depths aided by two complementary attention maps. This decomposition prevents depth estimation, which is a more difficult task, from interfering with the UV estimations at both the prediction and feature levels. The second innovation is PixDropout, which is, to the best of our knowledge, the first appearance-based data augmentation method for hand depth images. Experimental results demonstrate that the proposed model outperforms the state-of-the-art methods on three public benchmark datasets. Our implementation is available at https://github.com/mrezaei92/TriHorn-Net.
The 3D Average Precision (3D AP) relies on the intersection over union between predictions and ground truth objects. However, camera-only detectors have limited depth accuracy, which may cause otherwise reasonable predictions that suffer from such longitudinal localization errors to be treated as false positives. We therefore propose variants of the 3D AP metric to be more permissive with respect to depth estimation errors. Specifically, our novel longitudinal error tolerant metrics, LET-3D-AP and LET-3D-APL, allow longitudinal localization errors of the prediction boxes up to a given tolerance. To evaluate the proposed metrics, we also construct a new test set for the Waymo Open Dataset, tailored to camera-only 3D detection methods. Surprisingly, we find that state-of-the-art camera-based detectors can outperform popular LiDAR-based detectors with our new metrics past at 10% depth error tolerance, suggesting that existing camera-based detectors already have the potential to surpass LiDAR-based detectors in downstream applications. We believe the proposed metrics and the new benchmark dataset will facilitate advances in the field of camera-only 3D detection by providing more informative signals that can better indicate the system-level performance.
Monocular 3D object detection has attracted great attention for its advantages in simplicity and cost. Due to the ill-posed 2D to 3D mapping essence from the monocular imaging process, monocular 3D object detection suffers from inaccurate depth estimation and thus has poor 3D detection results. To alleviate this problem, we propose to introduce the ground plane as a prior in the monocular 3d object detection. The ground plane prior serves as an additional geometric condition to the ill-posed mapping and an extra source in depth estimation. In this way, we can get a more accurate depth estimation from the ground. Meanwhile, to take full advantage of the ground plane prior, we propose a depth-align training strategy and a precise two-stage depth inference method tailored for the ground plane prior. It is worth noting that the introduced ground plane prior requires no extra data sources like LiDAR, stereo images, and depth information. Extensive experiments on the KITTI benchmark show that our method could achieve state-of-the-art results compared with other methods while maintaining a very fast speed. Our code and models are available at https://github.com/cfzd/MonoGround.
A self-driving perception model aims to extract 3D semantic representations from multiple cameras collectively into the bird's-eye-view (BEV) coordinate frame of the ego car in order to ground downstream planner. Existing perception methods often rely on error-prone depth estimation of the whole scene or learning sparse virtual 3D representations without the target geometry structure, both of which remain limited in performance and/or capability. In this paper, we present a novel end-to-end architecture for ego 3D representation learning from an arbitrary number of unconstrained camera views. Inspired by the ray tracing principle, we design a polarized grid of "imaginary eyes" as the learnable ego 3D representation and formulate the learning process with the adaptive attention mechanism in conjunction with the 3D-to-2D projection. Critically, this formulation allows extracting rich 3D representation from 2D images without any depth supervision, and with the built-in geometry structure consistent w.r.t. BEV. Despite its simplicity and versatility, extensive experiments on standard BEV visual tasks (e.g., camera-based 3D object detection and BEV segmentation) show that our model outperforms all state-of-the-art alternatives significantly, with an extra advantage in computational efficiency from multi-task learning.
Self-supervised monocular depth estimation has been a subject of intense study in recent years, because of its applications in robotics and autonomous driving. Much of the recent work focuses on improving depth estimation by increasing architecture complexity. This paper shows that state-of-the-art performance can also be achieved by improving the learning process rather than increasing model complexity. More specifically, we propose (i) disregarding small potentially dynamic objects when training, and (ii) employing an appearance-based approach to separately estimate object pose for truly dynamic objects. We demonstrate that these simplifications reduce GPU memory usage by 29% and result in qualitatively and quantitatively improved depth maps. The code is available at https://github.com/kieran514/Dyna-DM.
Scene flow represents the motion of points in the 3D space, which is the counterpart of the optical flow that represents the motion of pixels in the 2D image. However, it is difficult to obtain the ground truth of scene flow in the real scenes, and recent studies are based on synthetic data for training. Therefore, how to train a scene flow network with unsupervised methods based on real-world data shows crucial significance. A novel unsupervised learning method for scene flow is proposed in this paper, which utilizes the images of two consecutive frames taken by monocular camera without the ground truth of scene flow for training. Our method realizes the goal that training scene flow network with real-world data, which bridges the gap between training data and test data and broadens the scope of available data for training. Unsupervised learning of scene flow in this paper mainly consists of two parts: (i) depth estimation and camera pose estimation, and (ii) scene flow estimation based on four different loss functions. Depth estimation and camera pose estimation obtain the depth maps and camera pose between two consecutive frames, which provide further information for the next scene flow estimation. After that, we used depth consistency loss, dynamic-static consistency loss, Chamfer loss, and Laplacian regularization loss to carry out unsupervised training of the scene flow network. To our knowledge, this is the first paper that realizes the unsupervised learning of 3D scene flow from monocular camera. The experiment results on KITTI show that our method for unsupervised learning of scene flow meets great performance compared to traditional methods Iterative Closest Point (ICP) and Fast Global Registration (FGR). The source code is available at: https://github.com/IRMVLab/3DUnMonoFlow.
This paper presents a unified framework for depth-aware panoptic segmentation (DPS), which aims to reconstruct 3D scene with instance-level semantics from one single image. Prior works address this problem by simply adding a dense depth regression head to panoptic segmentation (PS) networks, resulting in two independent task branches. This neglects the mutually-beneficial relations between these two tasks, thus failing to exploit handy instance-level semantic cues to boost depth accuracy while also producing sub-optimal depth maps. To overcome these limitations, we propose a unified framework for the DPS task by applying a dynamic convolution technique to both the PS and depth prediction tasks. Specifically, instead of predicting depth for all pixels at a time, we generate instance-specific kernels to predict depth and segmentation masks for each instance. Moreover, leveraging the instance-wise depth estimation scheme, we add additional instance-level depth cues to assist with supervising the depth learning via a new depth loss. Extensive experiments on Cityscapes-DPS and SemKITTI-DPS show the effectiveness and promise of our method. We hope our unified solution to DPS can lead a new paradigm in this area. Code is available at https://github.com/NaiyuGao/PanopticDepth.
This paper presents a learning-based method for multi-view depth estimation from posed images. Our core idea is a "learning-to-optimize" paradigm that iteratively indexes a plane-sweeping cost volume and regresses the depth map via a convolutional Gated Recurrent Unit (GRU). Since the cost volume plays a paramount role in encoding the multi-view geometry, we aim to improve its construction both at pixel- and frame- levels. At the pixel level, we propose to break the symmetry of the Siamese network (which is typically used in MVS to extract image features) by introducing a transformer block to the reference image (but not to the source images). Such an asymmetric volume allows the network to extract global features from the reference image to predict its depth map. Given potential inaccuracies in the poses between reference and source images, we propose to incorporate a residual pose network to correct the relative poses. This essentially rectifies the cost volume at the frame level. We conduct extensive experiments on real-world MVS datasets and show that our method achieves state-of-the-art performance in terms of both within-dataset evaluation and cross-dataset generalization. Code available: https://github.com/oppo-us-research/riav-mvs.
Masked image modeling (MIM) as pre-training is shown to be effective for numerous vision downstream tasks, but how and where MIM works remain unclear. In this paper, we compare MIM with the long-dominant supervised pre-trained models from two perspectives, the visualizations and the experiments, to uncover their key representational differences. From the visualizations, we find that MIM brings locality inductive bias to all layers of the trained models, but supervised models tend to focus locally at lower layers but more globally at higher layers. That may be the reason why MIM helps Vision Transformers that have a very large receptive field to optimize. Using MIM, the model can maintain a large diversity on attention heads in all layers. But for supervised models, the diversity on attention heads almost disappears from the last three layers and less diversity harms the fine-tuning performance. From the experiments, we find that MIM models can perform significantly better on geometric and motion tasks with weak semantics or fine-grained classification tasks, than their supervised counterparts. Without bells and whistles, a standard MIM pre-trained SwinV2-L could achieve state-of-the-art performance on pose estimation (78.9 AP on COCO test-dev and 78.0 AP on CrowdPose), depth estimation (0.287 RMSE on NYUv2 and 1.966 RMSE on KITTI), and video object tracking (70.7 SUC on LaSOT). For the semantic understanding datasets where the categories are sufficiently covered by the supervised pre-training, MIM models can still achieve highly competitive transfer performance. With a deeper understanding of MIM, we hope that our work can inspire new and solid research in this direction.
Convolutional Neural Networks (CNNs) are known for requiring extensive computational resources, and quantization is among the best and most common methods for compressing them. While aggressive quantization (i.e., less than 4-bits) performs well for classification, it may cause severe performance degradation in image-to-image tasks such as semantic segmentation and depth estimation. In this paper, we propose Wavelet Compressed Convolution (WCC) -- a novel approach for high-resolution activation maps compression integrated with point-wise convolutions, which are the main computational cost of modern architectures. To this end, we use an efficient and hardware-friendly Haar-wavelet transform, known for its effectiveness in image compression, and define the convolution on the compressed activation map. We experiment with various tasks that benefit from high-resolution input. By combining WCC with light quantization, we achieve compression rates equivalent to 1-4bit activation quantization with relatively small and much more graceful degradation in performance. Our code is available at https://github.com/BGUCompSci/WaveletCompressedConvolution.
Self-supervised monocular depth estimation has been widely studied recently. Most of the work has focused on improving performance on benchmark datasets, such as KITTI, but has offered a few experiments on generalization performance. In this paper, we investigate the backbone networks (e.g. CNNs, Transformers, and CNN-Transformer hybrid models) toward the generalization of monocular depth estimation. We first evaluate state-of-the-art models on diverse public datasets, which have never been seen during the network training. Next, we investigate the effects of texture-biased and shape-biased representations using the various texture-shifted datasets that we generated. We observe that Transformers exhibit a strong shape bias and CNNs do a strong texture-bias. We also find that shape-biased models show better generalization performance for monocular depth estimation compared to texture-biased models. Based on these observations, we newly design a CNN-Transformer hybrid network with a multi-level adaptive feature fusion module, called MonoFormer. The design intuition behind MonoFormer is to increase shape bias by employing Transformers while compensating for the weak locality bias of Transformers by adaptively fusing multi-level representations. Extensive experiments show that the proposed method achieves state-of-the-art performance with various public datasets. Our method also shows the best generalization ability among the competitive methods.