当前位置:网站首页>Robust 3D Object Detection in Cold Weather Conditions
Robust 3D Object Detection in Cold Weather Conditions
2022-08-11 06:16:00 【zhSunw】
Robust 3D Object Detection in Cold Weather Conditions
Method
Point Sampling from 3D Shape Reconstructions
- Use alpha shapes to reconstruct a 3D surface S from the exhaust condensation point cloud in the original point cloud
- Uniformly sample N∈[100,1000] points from surface S
- A new sample is obtained by assigning the reflection intensity to each point according to the nearest principle
Point Cloud Augmentation Strategy
- Probabilistically placed in clean point cloud data in generating a large number of exhaust gas condensation point cloud samples:
- Pgas probability is generated in the rear center, right rear corner or left rear corner of the target
- Ptop probability is generated at the top of the target
- Paug total probability controls the variation of noise in the data
- Convert each point to spherical coordinates, resample the point cloud using the same resolution as the sensor's parameter settings: solves an issue where applying point cloud augmentation would violate the dataset's sensor physics
Noise Robustness Loss
Introduce noise loss: IoU between the real frame of the exhaust gas condensation point cloud and the predicted target frame (reduce the noise in the predicted target frame - calculate the loss with the number of noise points in the frame:
Schematic of how it works:
Experiments
Comparison of vehicle classes on the DENSE test set:
Aug represents the model retrained using data augmentation and adding noise loss.Most of the test data does not include exhaust emissions, and other weather effects can side-effect the enhancement of the model.
Comparison of vehicle class accuracy on the DENSE-GAS test set:
Combined with a noise robustness loss, forces the network to learn to distinguish between vehicles and nearby noise points
Accuracy comparison on the DENSE test set following the TANet noise experiment:
Noise represents the amount of noise added in the target box
Ablation experiment:
PointRCNN's vehicle class accuracy on DENSE-GAS test set
- No noise loss is applicable, no matter what kind of data augmentation is not conducive to network training
- Using the proposed generation method is significantly better than adding random noise to the target box
The effect of noise loss weights on the model:
- Too low is not conducive to the model paying attention to noise
- Too high is easy to keep the model away from noise, but also the target position (only focus on noise)
边栏推荐
- 关于安全帽识别系统,你需要知道的选择要点
- 【sqlyog】【mysql】csv导入问题
- GBase 8a MPP Cluster产品高级特性
- Toolbar 和 DrawerLayout 滑动菜单
- 【mysql】查询不区分大小写(用户密码登录不区分大小写)
- 梅科尔工作室-HarmonyOS应用开发的第二次培训
- Rethinking LiDAR Object Detection in adverse weather conditions
- CVPR2020: Seeing Through Fog Without Seeing Fog
- Maykel Studio - Django Web Application Framework + MySQL Database Second Training
- The working principle and industry application of AI intelligent image recognition
猜你喜欢
随机推荐
Toolbar 和 DrawerLayout 滑动菜单
软件架构之--MVC、MVP、MVVM
基于AI智能图像识别:4个不同的行业应用
Use regex to verify whether the file name is legal
GBase 8s共享内存中的常驻内存段
AI智能图像识别的工作原理及行业应用
CVPR2022——A VERSATILE MULTI-VIEW FRAMEWORK
GBase 8s的多线程结构
mAPH - Waymo dataset
2021-05-10
OSPF综合实验
【sqlyog】【mysql】csv导入问题
XSS跨站脚本攻击详解以及复现gallerycms字符长度限制短域名绕过
GBase 8s的分片和索引
安全帽佩戴识别系统介绍
跳转到微信小程序方法
Androd 基本布局(其一)
Thread Handler
NodeRed系列—发送消息给emqx
Redis哨兵模式