Achieving Real-Time Object Detection on MobileDevices with Neural
Pruning Search
- URL: http://arxiv.org/abs/2106.14943v1
- Date: Mon, 28 Jun 2021 18:59:20 GMT
- Title: Achieving Real-Time Object Detection on MobileDevices with Neural
Pruning Search
- Authors: Pu Zhao, Wei Niu, Geng Yuan, Yuxuan Cai, Bin Ren, Yanzhi Wang, Xue Lin
- Abstract summary: We propose a compiler-aware neural pruning search framework to achieve high-speed inference on autonomous vehicles for 2D and 3D object detection.
For the first time, the proposed method achieves computation (close-to) real-time, 55ms and 99ms inference times for YOLOv4 based 2D object detection and PointPillars based 3D detection.
- Score: 45.20331644857981
- License: http://creativecommons.org/licenses/by-sa/4.0/
- Abstract: Object detection plays an important role in self-driving cars for security
development. However, mobile systems on self-driving cars with limited
computation resources lead to difficulties for object detection. To facilitate
this, we propose a compiler-aware neural pruning search framework to achieve
high-speed inference on autonomous vehicles for 2D and 3D object detection. The
framework automatically searches the pruning scheme and rate for each layer to
find a best-suited pruning for optimizing detection accuracy and speed
performance under compiler optimization. Our experiments demonstrate that for
the first time, the proposed method achieves (close-to) real-time, 55ms and
99ms inference times for YOLOv4 based 2D object detection and PointPillars
based 3D detection, respectively, on an off-the-shelf mobile phone with minor
(or no) accuracy loss.
Related papers
- oTTC: Object Time-to-Contact for Motion Estimation in Autonomous Driving [4.707950656037167]
autonomous driving systems rely heavily on object detection to avoid collisions and drive safely.
Monocular 3D object detectors try to solve this problem by directly predicting 3D bounding boxes and object velocities given a camera image.
Recent research estimates time-to-contact in a per-pixel manner and suggests that it is more effective measure than velocity and depth combined.
We propose per-object time-to-contact estimation by extending object detection models to additionally predict the time-to-contact attribute for each object.
arXiv Detail & Related papers (2024-05-13T12:34:18Z) - Unsupervised Domain Adaptation for Self-Driving from Past Traversal
Features [69.47588461101925]
We propose a method to adapt 3D object detectors to new driving environments.
Our approach enhances LiDAR-based detection models using spatial quantized historical features.
Experiments on real-world datasets demonstrate significant improvements.
arXiv Detail & Related papers (2023-09-21T15:00:31Z) - CFTrack: Center-based Radar and Camera Fusion for 3D Multi-Object
Tracking [9.62721286522053]
We propose an end-to-end network for joint object detection and tracking based on radar and camera sensor fusion.
Our proposed method uses a center-based radar-camera fusion algorithm for object detection and utilizes a greedy algorithm for object association.
We evaluate our method on the challenging nuScenes dataset, where it achieves 20.0 AMOTA and outperforms all vision-based 3D tracking methods in the benchmark.
arXiv Detail & Related papers (2021-07-11T23:56:53Z) - 2nd Place Solution for Waymo Open Dataset Challenge - Real-time 2D
Object Detection [26.086623067939605]
In this report, we introduce a real-time method to detect the 2D objects from images.
We leverage accelerationRT to optimize the inference time of our detection pipeline.
Our framework achieves the latency of 45.8ms/frame on an Nvidia Tesla V100 GPU.
arXiv Detail & Related papers (2021-06-16T11:32:03Z) - Analysis of voxel-based 3D object detection methods efficiency for
real-time embedded systems [93.73198973454944]
Two popular voxel-based 3D object detection methods are studied in this paper.
Our experiments show that these methods mostly fail to detect distant small objects due to the sparsity of the input point clouds at large distances.
Our findings suggest that a considerable part of the computations of existing methods is focused on locations of the scene that do not contribute with successful detection.
arXiv Detail & Related papers (2021-05-21T12:40:59Z) - Exploiting Playbacks in Unsupervised Domain Adaptation for 3D Object
Detection [55.12894776039135]
State-of-the-art 3D object detectors, based on deep learning, have shown promising accuracy but are prone to over-fit to domain idiosyncrasies.
We propose a novel learning approach that drastically reduces this gap by fine-tuning the detector on pseudo-labels in the target domain.
We show, on five autonomous driving datasets, that fine-tuning the detector on these pseudo-labels substantially reduces the domain gap to new driving environments.
arXiv Detail & Related papers (2021-03-26T01:18:11Z) - Achieving Real-Time LiDAR 3D Object Detection on a Mobile Device [53.323878851563414]
We propose a compiler-aware unified framework incorporating network enhancement and pruning search with the reinforcement learning techniques.
Specifically, a generator Recurrent Neural Network (RNN) is employed to provide the unified scheme for both network enhancement and pruning search automatically.
The proposed framework achieves real-time 3D object detection on mobile devices with competitive detection performance.
arXiv Detail & Related papers (2020-12-26T19:41:15Z) - Kinematic 3D Object Detection in Monocular Video [123.7119180923524]
We propose a novel method for monocular video-based 3D object detection which carefully leverages kinematic motion to improve precision of 3D localization.
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.
arXiv Detail & Related papers (2020-07-19T01:15:12Z)
This list is automatically generated from the titles and abstracts of the papers in this site.
This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.