Difficulty-Aware Label-Guided Denoising for Monocular 3D Object Detection

Perceiving the physical world in 3D is fundamental for self-driving applications. Although temporal motion is an invaluable resource to human vision for detection, tracking, and depth perception, such features have not been thoroughly utilized in modern 3D object detectors. In this work, we propose a novel method for monocular video-based 3D object detection which leverages kinematic motion to extract scene dynamics and improve localization accuracy. We first propose a novel decomposition of object orientation and a self-balancing 3D confidence. We show that both components are critical to enable our kinematic model to work effectively. Collectively, using only a single model, we efficiently leverage 3D kinematics from monocular videos to improve the overall localization precision in 3D object detection while also producing useful by-products of scene dynamics (ego-motion and per-object velocity). We achieve state-of-the-art performance on monocular 3D object detection and the Bird’s Eye View tasks within the KITTI self-driving dataset.Keywords3D object detectionMonocularVideoPhysics-based

Monocular 3D object detection has been studied as a promising task for diverse applications, such as autonomous driving, due to its lower cost and more straightforward configuration than multiple sensors. However, existing studies have focused on clear weather without considering diverse weather conditions with varying intensity, such as rain, snow, and fog, affecting detection performance. In this paper, we propose MonoRange, a monocular 3D object detection method that uses object-centric images and range maps in adverse weather conditions. Leveraging the 2D detection results, MonoRange generates range maps from images via an object-centric range map reconstruction. Furthermore, MonoRange flexibly removes adverse weather noise in images via weather intensity adaptive image restoration with a weight modulation transformer. Then, MonoRange fuses the range map and restored image and predicts 3D bounding boxes via the range map aligned detector. Introducing the projected box consistency loss between 2D and 3D boxes also enables to consistent and accurate 3D object detection. Experimental results on diverse weather datasets demonstrate that MonoRange surpasses existing monocular 3D object detection approaches. The source code is available at <uri xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">https://github.com/jhyoon964/MonoRange</uri>

Probabilistic instance shape reconstruction with sparse LiDAR for monocular 3D object detection

Monocular 3D object detection attempts to predict object location and dimension in 3D space using a single optical camera. The main challenge of monocular 3D object detection is the lack of depth information to infer the object&#x2019;s distance. In this work, we propose a monocular 3D object detection based on CenterNet with discrete depth and encoded orientation angle. Our proposed method is able to achieve 54.6% detection score for car class on the challenging Cityscapes autonomous driving dataset, outperforming prior monocular 3D object detection by a convincing margin.

Perceiving 3D objects from monocular inputs is crucial for robotic systems, given its economy compared to multi-sensor settings. It is notably difficult as a single image can not provide any clues for predicting absolute depth values. Motivated by binocular methods for 3D object detection, we take advantage of the strong geometry structure provided by camera ego-motion for accurate object depth estimation and detection. We first make a theoretical analysis on this general two-view case and notice two challenges: 1) Cumulative errors from multiple estimations that make the direct prediction intractable; 2) Inherent dilemmas caused by static cameras and matching ambiguity. Accordingly, we establish the stereo correspondence with a geometry-aware cost volume as the alternative for depth estimation and further compensate it with monocular understanding to address the second problem. Our framework, named Depth from Motion (DfM), then uses the established geometry to lift 2D image features to the 3D space and detects 3D objects thereon. We also present a pose-free DfM to make it usable when the camera pose is unavailable. Our framework outperforms state-of-the-art methods by a large margin on the KITTI benchmark. Detailed quantitative and qualitative analyses also validate our theoretical conclusions. The code is released at https://github.com/Tai-Wang/Depth-from-Motion. KeywordsMonocular 3D object detectionDepth from motion

Monocular 3D object detection using dual quadric for autonomous driving

Monocular 3D object detection is an essential task for infrastructure-less autonomous navigation and driving due to its low cost. Most previous state-of-the-art monocular 3D object detection methods depended on Convolutional Neural Networks (CNNs). We show that this reliance on CNNs is not necessary and a Transformer-based method can also perform very well. In this paper, we present the development of a new and general framework DST3D to predict a 3D bounding box for each object based on DLA-Swin Transformer (DST) in an end-to-end fashion without any pre-trained network for depth estimation. We propose an object-scale adaptive Gaussian Kernel for generating ground truth keypoint heatmap, which associate the keypoint of the object with its size and help to enhance the network’s performance. In addition, while regressing 3D variables, we introduce a double predictions dropout loss, which significantly improves both training consistence and detection accuracy. All of these make our framework simple yet efficient. Compared with all state-of-the-art CNNs-based methods, our proposed DST3D network achieves comparative performance on challenging KITTI benchmark with a faster speed, giving the top results on both 3D object detection and Bird’s Eye View evaluation.

The objective of this paper is to learn context- and depth- aware feature representation to solve the problem of monocular 3D object detection. We make following contributions: (i) rather than appealing to the complicated pseudo-LiDAR based approach, we propose a depth-conditioned dynamic message propagation (DDMP) network to effectively integrate the multi-scale depth information with the image context; (ii) this is achieved by first adaptively sampling context-aware nodes in the image context and then dynamically predicting hybrid depth-dependent filter weights and affinity matrices for propagating information; (Hi) by augmenting a center-aware depth encoding (CDE) task, our method successfully alleviates the inaccurate depth prior; (iv) we thoroughly demonstrate the effectiveness of our proposed approach and show state-of-the-art results among the monocular-based approaches on the KITTI benchmark dataset. Particularly, we rank 1 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">st</sup> in the highly competitive KITTI monocular 3D object detection track on the submission day (November 16th, 2020). Code and models are released at https: //github.com/fudan-zvg/DDMP

Background: 3D object detection based on point clouds in road scenes has attracted much attention recently. The voxel-based methods voxelize the scene to regular grids, which can be processed with the advanced feature learning frameworks based on convolutional layers for semantic feature learning. The point-based methods can extract the geometric feature of the point due to the coordinate reservations. The combination of the two is effective for 3D object detection. However, the current methods use a voxel-based detection head with anchors for classification and localization. Although the preset anchors cover the entire scene, it is not suitable for detection tasks with larger scenes and multiple categories of objects, due to the limitation of the voxel size. Additionally, the misalignment between the predicted confidence and proposals in the Regions of the Interest (ROI) selection bring obstacles to 3D object detection. Methods: We investigate the combination of voxel-based methods and point-based methods for 3D object detection. Additionally, a voxel-to-point module that captures semantic and geometric features is proposed in the paper. The voxel-to-point module is conducive to the detection of small-size objects and avoids the presets of anchors in the inference stage. Moreover, a confidence adjustment module with the center-boundary-aware confidence attention is proposed to solve the misalignment between the predicted confidence and proposals in the regions of the interest selection. Results: The proposed method has achieved state-of-the-art results for 3D object detection in the Karlsruhe Institute of Technology and Toyota Technological Institute (KITTI) object detection dataset. Actually, as of September 19, 2021, our method ranked 1st in the 3D and Bird Eyes View (BEV) detection of cyclists tagged with difficulty level ‘easy’, and ranked 2nd in the 3D detection of cyclists tagged with ‘moderate’. Conclusions: We propose an end-to-end two-stage 3D object detector with voxel-to-point module and confidence adjustment module.

Advanced driver assistance systems (ADAS) mainly consist of three components: environmental perception, decision planning, and motion control. As a fundamental component of the ADAS environmental perception system, 3D object detection enables vehicles to avoid obstacles and ensure driving safety only through accurate and real‐time prediction and localization of three‐dimensional targets such as vehicles and pedestrians in road scenes. Therefore, to improve both the real‐time performance and accuracy of 3D object detection, we propose a lightweight depth prediction‐based 3D object detection model with multi‐scale fusion—SPWS‐Transformer. First, to enhance the model's accuracy, we propose a feature extraction network incorporating multi‐scale feature fusion and depth prediction. By designing a multi‐scale feature fusion module, we effectively combine multi‐scale semantic and fine‐grained information from feature maps of different scales to enhance the network's feature extraction capability. To capture spatial information from the feature maps, we apply convolution, group normalization, and nonlinear activation operations on the fused feature maps to generate depth feature maps. Both the fused feature maps and depth feature maps serve as inputs for subsequent network stages. To further improve accuracy, we leverage the long‐range modelling advantages of Transformers by designing a feature enhancement encoder to strengthen the representation capability of depth feature maps. We incorporate a dilated encoder to perform positional encoding on depth feature maps and utilize multi‐head self‐attention mechanisms to capture contextual relationships within the input scene, thereby enhancing the detection capability of the 3D object detection network. Then, to improve real‐time performance, we design a decoder structure with scale‐aware attention. By predefining masks of different scales, we adaptively learn a scale‐aware filter using depth and visual features to enhance object queries. Finally, on the KITTI dataset, the improved algorithm achieves an AP of 24.66% for the car category, with more significant improvements in detection accuracy under the ‘hard’ difficulty level. The model achieves an inference time of 24 ms.

Monocular 3D object detection is an important yet challenging task in autonomous driving. Some existing methods leverage depth information from an off-the-shelf depth estimator to assist 3D detection, but suffer from the additional computational burden and achieve limited performance caused by inaccurate depth priors. To alleviate this, we propose MonoDTR, a novel end-to-end depth-aware transformer network for monocular 3D object detection. It mainly consists of two components: (1) the Depth-Aware Feature Enhancement (DFE) module that implicitly learns depth-aware features with auxiliary supervision without requiring extra computation, and (2) the Depth-Aware Transformer (DTR) module that globally integrates context- and depth-aware features. Moreover, different from conventional pixel-wise positional encodings, we introduce a novel depth positional encoding (DPE) to inject depth positional hints into transformers. Our proposed depth-aware modules can be easily plugged into existing image-only monocular 3D object detectors to improve the performance. Extensive experiments on the KITTI dataset demonstrate that our approach outperforms previous state-of-the-art monocular-based methods and achieves real-time detection. Code is available at https://github.com/kuanchihhuang/.MonoDTR.

Point cloud-based three-dimensional (3D) object detection is a critical task in autonomous driving, robotics, and augmented reality, where accurate localization and classification of objects are essential under diverse and challenging scenarios. This work introduces a point cloud-based multi-target 3D object detection framework using Light Detection and Ranging (LiDAR) sensors. The key contribution lies in improving the encoding technique to preserve spatial information and adapt to varying point densities, enabling efficient processing of raw LiDAR data for robust 3D detection. Specifically, we design an efficient network architecture that incorporates a single unified classification head to jointly handle positive and negative samples, simplifying the design and improving training stability. Our model detects all classes (car, pedestrian, cyclist) simultaneously within a single network, enhancing computational efficiency while avoiding separate networks for each class. Furthermore, the proposed feature extraction pipeline combines a voxel-based encoder with a sparsely embedded convolutional backbone and a feature pyramid network, facilitating multi-scale feature representation and effective detection of objects at different scales. This encoding method facilitates the efficient processing of raw LiDAR data, enabling accurate object detection across diverse scenarios. As a result, our model achieves state-of-the-art performance on the KITTI dataset, surpassing baseline methods such as PointPillars and VoxelNet by delivering superior Average Precision (AP) across all difficulty levels for the car class in both bird’s-eye view (BEV) and 3D object detection tasks. Specifically, it achieves AP40 scores of 92.84%, 88.63%, and 85.68% in BEV detection and 87.84%, 76.65%, and 73.60% in 3D detection for Easy, Moderate, and Hard levels, respectively. These results highlight the effectiveness of our encoding technique in enhancing model efficiency and detection accuracy across diverse scenarios.

Monocular 3D object detection has recently shown promising results, however there remain challenging problems. One of those is the lack of invariance to different camera intrinsic parameters, which can be observed across different 3D object datasets. Little effort has been made to exploit the combination of heterogeneous 3D object datasets. In contrast to general intuition, we show that more data does not automatically guarantee a better performance, but rather, methods need to have a degree of 'camera independence' in order to benefit from large and heterogeneous training data. In this paper we propose a category-level pose estimation method based on instance segmentation, using camera independent geometric reasoning to cope with the varying camera viewpoints and intrinsics of different datasets. Every pixel of an instance predicts the object dimensions, the 3D object reference points projected in 2D image space and, optionally, the local viewing angle. Camera intrinsics are only used outside of the learned network to lift the predicted 2D reference points to 3D. We surpass camera independent methods on the challenging KITTI3D benchmark and show the key benefits compared to camera dependent methods.

Monocular 3D object detection, with the aim of predicting the geometric properties of on-road objects, is a promising research topic for the intelligent perception systems of autonomous driving. Most state-of-the-art methods follow a keypoint-based paradigm, where the keypoints of objects are predicted and employed as the basis for regressing the other geometric properties. In this work, a unified network named as FADNet is presented to address the task of monocular 3D object detection. In contrast to previous keypoint-based methods, we propose to divide the output modalities into different groups according to the estimation difficulty of object properties. Different groups are treated differently and sequentially associated by a convolutional Gated Recurrent Unit. Another contribution of this work is the strategy of depth hint augmentation. To provide characterized depth patterns as hints for depth estimation, a dedicated depth hint module is designed to generate row-wise features named as depth hints, which are explicitly supervised in a bin-wise manner. The contributions of this work are validated by conducting experiments and ablation study on the KITTI benchmark. Without utilizing depth priors, post optimization, or other refinement modules, our network performs competitively against state-of-the-art methods while maintaining a decent running speed.

Monocular 3D object detection is an essential component in autonomous driving while challenging to solve, especially for those occluded samples which are only partially visible. Most detectors consider each 3D object as an independent training target, inevitably resulting in a lack of useful information for occluded samples. To this end, we propose a novel method to improve the monocular 3D object detection by considering the relationship of paired samples. This allows us to encode spatial constraints for partially-occluded objects from their adjacent neighbors. Specifically, the proposed detector computes uncertainty-aware predictions for object locations and 3D distances for the adjacent object pairs, which are subsequently jointly optimized by nonlinear least squares. Finally, the one-stage uncertainty-aware prediction structure and the post-optimization module are dedicatedly integrated for ensuring the run-time efficiency. Experiments demonstrate that our method yields the best performance on KITTI 3D detection benchmark, by outperforming state-of-the-art competitors by wide margins, especially for the hard samples.