ESGN: Efficient Stereo Geometry Network for Fast 3D Object Detection

Gao, Aqi; Pang, Yanwei; Nie, Jing; Shao, Zhuang; Cao, Jiale; Guo, Yishun; Li, Xuelong

doi:10.1109/tcsvt.2022.3202810

Cited by 18 publications

(7 citation statements)

References 58 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Besides this leadership in inference time, ER3D also outperformed RTS3D [12], YOLOStereo3D [13] and SCNetc [23] with notable leadership in accuracy. Although ESGN [4] achieves slightly higher accuracy than ER3D, it is worth noting that ER3D achieves 3 × faster inference speed than ESGN [4] using the GPU with half the computing power than ESGN [4].…”

Section: B Resultsmentioning

confidence: 99%

“…Comparison with State-of-the-art Methods. In this evaluation, we compare ER3D with state-of-the-art approaches that can be divided into accuracy-first approaches [2], [6], [11], [17], [18], [25], [29] and efficiency-first approaches [4], [12], [13], [23], [26]. Tab.…”

Section: B Resultsmentioning

confidence: 99%

“…Differently from the above approaches. Gao et al [4] down-scales the feature generalization to speed up the algorithmic latency. Wang et al [26] conduct the replacement of heavy cost volume with lightweight Pseudl-LiDAR feature volume to gain the advance in computing latency.…”

Section: B Stereo-based 3d Object Detectionmentioning

confidence: 99%

“…Generally, these methods can be broadly categorized into accuracy-first approaches and efficiency-first approaches according to the emphasis aspects. The advancements that prioritize detection accuracy [2], [4], [17] typically propose a diverse array of novel architectures and system pipelines with the aim of elaborating the undiscovered optimization opportunities. Although these proposals tend to be accurate and reliable in producing 3D detection results, the bulky and high computational complexity architecture hinders the deployment of resource-constrained embedded systems (e.g., drones).…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

ER3D: An Efficient Real-time 3D Object Detection Framework for Autonomous Driving

Meng,

Li,

Chen

et al. 2023

2023 IEEE 29th International Conference on Parallel and Distributed Systems (ICPADS)

View full text Add to dashboard Cite

3D object detection is a vital computer vision task in mobile robotics and autonomous driving. However, most existing methods have exclusively focused on achieving high accuracy, leading to complex and bulky systems that can not be deployed in a real-time manner. In this paper, we propose the ER3D (Efficient and Real-time 3D) object detection framework, which takes stereo images as input and predicts 3D bounding boxes. Instead of using the complex network architecture, we leverage a fastbut-inaccurate method semi-global matching (SGM) for depth estimation. To eliminate the accuracy degradation in 3D detection caused by inaccurate depth estimation, we introduce decoupled regression head and 3D distance-consistency IoU loss to boost the accuracy performance of the 3D detector with a small computing overhead. ER3D achieves both high-precision and realtime performance to enable practical applications of 3D object detection systems on robotic systems. Extensive experiments with the comparison of the state of the arts demonstrate the superior practicability of ER3D, which achieves comparable detection accuracy with significant leadership on inference efficiency.

show abstract

Section: B Resultsmentioning

confidence: 99%

Section: B Resultsmentioning

confidence: 99%

Section: B Stereo-based 3d Object Detectionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

ER3D: An Efficient Real-time 3D Object Detection Framework for Autonomous Driving

Meng,

Li,

Chen

et al. 2023

2023 IEEE 29th International Conference on Parallel and Distributed Systems (ICPADS)

View full text Add to dashboard Cite

show abstract

“…They use the anchor-free 2D box association method by detecting only objects in the left images and computing the left-right associations by predicting the distance between them. Gao et al [87] put forth an efficient stereo geometry network (ESGN) for 3D object detection. The ResNet-34 [88] backbone is used to extract multi-scale feature maps.…”

Section: B Stereo Images Methodsmentioning

confidence: 99%

Deep Learning-Based Image 3-D Object Detection for Autonomous Driving: Review

Alaba

Ball

2023

IEEE Sensors J.

View full text Add to dashboard Cite

An accurate and robust perception system is key to understanding the driving environment of autonomous driving and robots. Autonomous driving needs 3D information about objects, including the object's location and pose, to understand the driving environment clearly. A camera sensor is widely used in autonomous driving because of its richness in color and texture and low price. The major problem with the camera is the lack of 3D information, which is necessary to understand the 3D driving environment. Additionally, the object's scale change and occlusion make 3D object detection more challenging. Many deep learning-based methods, such as depth estimation, have been developed to solve the lack of 3D information. This survey presents the image 3D object detection 3D bounding box encoding techniques and evaluation metrics. The image-based methods are categorized based on the technique used to estimate an image's depth information, and insights are added to each method. Then, state-of-the-art (SOTA) monocular and stereo camera-based methods are summarized. We also compare the performance of the selected 3D object detection models and present challenges and future directions in 3D object detection.

show abstract

View-target relation-guided unsupervised 2D image-based 3D model retrieval via transformer

Chang,

Zhang,

Shao

2023

Multimedia Systems

View full text Add to dashboard Cite

Unsupervised 2D image-based 3D model retrieval aims at retrieving images from the gallery of 3D models by the given 2D images. Despite the encouraging progress made in this task, there are still two significant limitations: (1) feature alignment of 2D images and 3D model gallery is still difficult due to the huge gap between the two modalities. (2) The important view information in the 3D model gallery was ignored by the prior arts, which led to inaccurate results. To alleviate these limitations, inspired by the success of vision transformers (ViT) in a great variety of vision tasks, in this paper, we propose an end-to-end 3D model retrieval architecture on top of ViT, termly transformer-based 3D model retrieval network (T3DRN). In addition, to take advantage of the valuable view information of 3D models, we present an attentive module in T3DRN named shared view-guided attentive module (SVAM) to guide the learning of the alignment features. The proposed method is tested on the challenging dataset, MI3DOR-1. The extensive experimental results have proved the superiority of our proposed method to state-of-the-art methods.

show abstract

ESGN: Efficient Stereo Geometry Network for Fast 3D Object Detection

Abstract: How to cite:Please refer to published version for the most recent bibliographic citation information. If a published version is known of, the repository item page linked to above, will contain details on accessing it.

Cited by 18 publications

References 58 publications

ER3D: An Efficient Real-time 3D Object Detection Framework for Autonomous Driving

ER3D: An Efficient Real-time 3D Object Detection Framework for Autonomous Driving

Deep Learning-Based Image 3-D Object Detection for Autonomous Driving: Review

View-target relation-guided unsupervised 2D image-based 3D model retrieval via transformer

Contact Info

Product

Resources

About