RADIal is available now! Check the download section

valeo.ai

Last update: Jan 3, 2023

Related tags

Deep Learning RADIal

Overview

Latest news:

RADIal is available now! Check the download section. However, because we are currently working on the data anonymization, we provide for now a low resolution preview video stream. The full resolution will be provided once the anonymization is completed, planned by 2022, February.

RADIal dataset

RADIal stands for “Radar, Lidar et al.” It's a collection of 2-hour of raw data from synchronized automotive-grade sensors (camera, laser, High Definition radar) in various environments (citystreet, highway, countryside road) and comes with GPS and vehicle’s CAN traces.

RADIal contains 91 sequences of 1 to 4 minutes in duration, for a total of 2 hours. These sequences are categorized in highway, country-side and city driving. The distribution of the sequences is indicated in the figure below. Each sequence contains raw sensor signals recorded with their native frame rate. There are approximately 25,000 frames with the three sensors synchronized, out of which 8,252 are labelled with a total of 9,550 vehicles.

Sensor specifications

Central to the RADIal dataset, our high-definition radar is composed of NRx=16 receiving antennas and NTx= 12 transmitting antennas, leading to NRx·NTx= 192 virtual antennas. This virtual-antenna array enables reaching a high azimuth angular resolution while estimating objects’ elevation angles as well. As the radar signal is difficult to interpret by annotators and practitioners alike, a 16-layer automotive-grade laser scanner (LiDAR) and a 5 Mpix RGB camera are also provided. The camera is placed below the interior mirror behind the windshield while the radar and the LiDAR are installed in the middle of the front ventilation grid, one above the other. The three sensors have parallel horizontallines of sight, pointing in the driving direction. Their extrinsic parameters are provided together with the dataset. RADIal also offers synchronized GPS and CAN traces which give access to the geo-referenced position of the vehicle as well as its driving information such as speed, steering wheelangle and yaw rate. The sensors’ specifications are detailed in the table below.

Dataset structure

RADIal is a unique folder containing all the recorded sequences. Each sequence is a folder containing:

A preview video of the scene (low resolution);
The camera data compressed in MJPEG format (will be released by 2022, February);
The Laser Scanner point cloud data saved in a binary file;
The ADC radar data saved in a binary file. There are 4 files in total, one file for each radar chip, each chip containing 4 Rx antennas;
The GPS data saved in ASCII format
The CAN traces of the vehicle saved in binary format
And finally, a log file that provides the timestamp of each individual sensor event.

We provide in a Python library DBReader to read the data. Because all the radar data are recorded in a RAW format, that is to say the signal after the Analog to Digital Conversion (ADC), we provided too an optimized Python library SignalProcessing to process the Radar signal and generate either the Power Spectrums, the Point Cloud or the Range-Azimuth map.

Labels

Out of the 25,000 synchronized frames, 8,252 frames are labelled. Labels for vehicles are stored in a separated csv file. Each label containg the following information:

numSample: number of the current synchronized sample between all the sensors. That is to say, this label can be projected in each individual sensor with a common dataset_index value. Note that there might be more than one line with the same numSample, one line per label;
[x1_pix, y1_pix, x2_pix, y2_pix]: 2D coordinates of the vehicle' bouding boxes in the camera coordinates;
[laser_X_m, laser_Y_m, laser_Z_m]: 3D coordinates of the vehicle in the laser scanner coordinates system. Note that this 3D point is the middle of either the back or front visible face of the vehicle;
[radar_X_m, radar_Y_m, radar_R_m, radar_A_deg, radar_D, radar_P_db]: 2D coordinates (bird' eyes view) of the vehicle in the radar coordinates system either in cartesian (X,Y) or polar (R,A) coordinates. radar_D is the Doppler value and radar_P_db is the power of the reflected signal;
dataset: name of sequence it belongs to;
dataset_index: frame index in the current sequence;
Difficult: either 0 or 1

Note that -1 in all field means a frame without any label.

Labels for the Free-driving-space is provided as a segmentaion mask saved in a png file.

Download instructions

To download the raw dataset, please follow these instructions.

$ wget -c -i download_urls.txt -P your_target_path
$ unzip 'your_target_path/*.zip' -d your_target_path
$ rm -Rf your_target_path/*.zip

You will have then to use the SignalProcessing library to generate data for each modalities uppon your need.

We provide too a "ready to use" dataset that can be loaded with the PyTorch data loader example provided in the Loader folder.

$ wget https://www.dropbox.com/s/bvbndch5rucyp97/RADIal.zip

Comments

Radar details

Excellent dataset and nice code, thank you.

Could you provide more information about the radar system you are using? I would like to do some radar signal processing before applying classification algorithms. Specifically, I would need to know how the antenna arrays are placed (the relative distance between them in wavelengths).

opened by IgnacioRoldan 5
Raw Data Preprocessing

Hi, thank you for sharing the code! It is a really great job! I've been following your work recently and found that you've provided raw dataset and some preprocessing code in SignalProcessing folder. But I failed to preprocess the raw data into the "ready to use" format that you provided in https://drive.google.com/file/d/1PpAcL5r2PRYMxDb46ASps9YqToL7uJcE/view I' m wondering if you can provide the complete code about how to process the raw data into the "ready to use" dataset ? Thanks a lot!

opened by illaMcbender 4
problem of downing dataset

Thanks for your great work!

Can you provide RADIal part dataset in the form of BaiduPan, because the files always are interrupted after one hour.

Thanks!

opened by sutiankang 3
Unexpected keyword argument 'offset_radar' of DBReader

Hello, When runing signalprocessing/radar_processing, I encountered the following error

Could you please explain the meaning of "offset_radar" and "offset_scala" ?

opened by zhai-EE 2
Is hard cases trained with the same configuration as the pre-trained model?

The pre-trained model that you provide in the link is for easy cases, right? Did you train the hard cases with the same configuration as the pre-trained model, only thing to adjust is setting the dataset loading 'difficult=True'?

opened by FrkWang 2
Problem to use the odometry

Hallo, Thank you for sharing the dataset! When I used example_sync_reader.ipynb to check the data, I met the following issue. (P.S. I used the same measurement as the original file : RECORD@2020-11-22_12.28.47 ) Any idea how to solve it?

BR Yi

opened by chasedreams-jy 2
radar parameters

Hello,could you please provide the initial parameters of the radar, such as bandwidth, chirp time, starting frequency, etc., because I want to further process the data, thank you

opened by lightzwc 2
Prepared dataset radar doppler clarification

Hello, When looking at the prepared dataset, specifically at the doppler data in radar_PCL, I can not find an explanation on how to interpret this data and can not find an example of usage of this specific data. This is for example not used in the loader/example_dataset.ipynb file. In the loader/dataset.py file there is a definition of the radar_PCL data as "range,azimuth,elevation,power,doppler,x,y,z,v". The doppler value is an 8-bit value associated with each datapoint. How is the doppler value to be interpreted and is there an example which I have missed?

opened by samberngit 2
signal process code part: 404 not found

Hi there, thanks for your work. this is really a great project for sensor fusion area. I wonder when can we have the radar signal process code as it shows 404 right now. Thank you so much.

opened by yyxr75 2
Interpreting radar RA map dimensions

Dear authors,

Thanks for sharing such an interesting dataset. I am trying to playing with it. However, I am having trouble understanding the RA map -- what is the size of each range bin from the RA map, and size of the azimuth bin?

I generated the RA map from the SignalProcessing toolkit. When examine the RA map, should I expect the provied labels (from labels.csv) overlay with the bright spots in the RA map? I expect so, but it is not the case for the sample I checked -- there might be something wrong with the way I interpret the RA bins though.

Thanks in advance!

opened by laixuezhongwen 1
Downloading Error Issue

I have downloaded RADIal.zip from location "https://drive.google.com/drive/folders/1RFRoJJd9ghfjRHjA8t_POwGzYdoUA1IY".

Even after getting completly downloaded, the extraction error thrown in the above snapshot. Kindly look into the above issue.

opened by ChidanandKumarKS 1
Calibration problem for 3D bbox annotation

Hi,

First of all, thank you for your great contribution.

I'm trying 3D annotation for 3D Object detection with Radial dataset.

However, when projecting the 3D bbox annotated in 3D lidar space onto the image plane, the calibration is hardly correct.

Can you give me some tips or exact internal/external calibration values?

opened by shawrby 0
Issues about the extrinsic parameter of sensors in the "ready to use" dataset

Could you please provide more detailed extrinsic parameter of sensors in the "ready to use" dataset so that we can better calibrate them?

Thanks a lot!

opened by lbhwyy 0
Are you using a 4D millimeter wave radar?

I can see in Figure 1 (a) of your published paper that the RADAR point cloud projected into the image has height information, but most of the existing vehicle borne 3D millimeter wave radars do not have height measurement capability, so I want to ask if you use 4D millimeter wave radar?

opened by chenhengwei1999 0
Question about evaluation

Thanks for your great work!

I have a problem about run_evaluation function and run_Full_evaluation function. When I use the same weight and dataloader to evaluation, I found that the metrics about mAP, mAR and mIoU are quite different. Can you tell me what is the problem about it.

Thanks!

opened by sutiankang 0
Problem of distributed training

Thanks for the great paper, dataset and code!

I tried to train the model with ready data using single GPU, it took roughly half day. So I tried to add some distributed training component, the training time decreased, but also the AP/AR/IOU values. Have you tested distributed training? How do you correctly set the parameters to ensure shorter training time and proper AP/AR/IOU values?

Thank you!

opened by eagles1812 2
Raw-data extraction

Hi,

I'm trying to recreate the ready to use data set from raw data. In the script that is shared in an earlier issue, a so called label_candidates_0deg.npy is loaded.

What is the function of this file and what does this look like. Or can someone explain te me how I can recreate the data using the signalProcessing library in this project?

Thanks in advance!

opened by JeroenBax 0

RADIal is available now! Check the download section

Related tags

Overview

Latest news:

RADIal dataset

Sensor specifications

Dataset structure

Labels

Download instructions

Comments

Owner

valeo.ai

Mahadi-Now - This Is Pakistani Just Now Login Tools

DR-GAN: Automatic Radial Distortion Rectification Using Conditional GAN in Real-Time

Exposure Time Calculator (ETC) and radial velocity precision estimator for the Near InfraRed Planet Searcher (NIRPS) spectrograph

We have implemented shaDow-GNN as a general and powerful pipeline for graph representation learning. For more details, please find our paper titled Deep Graph Neural Networks with Shallow Subgraph Samplers, available on arXiv (https//arxiv.org/abs/2012.01380).

ERISHA is a mulitilingual multispeaker expressive speech synthesis framework. It can transfer the expressivity to the speaker's voice for which no expressive speech corpus is available.

A public available dataset for road boundary detection in aerial images

This repository contains the implementations related to the experiments of a set of publicly available datasets that are used in the time series forecasting research space.

Repository for publicly available deep learning models developed in Rosetta community

A pytorch implementation of Detectron. Both training from scratch and inferring directly from pretrained Detectron weights are available.

PyTorch Implementation of Fully Convolutional Networks. (Training code to reproduce the original result is available.)

Transformers4Rec is a flexible and efficient library for sequential and session-based recommendation, available for both PyTorch and Tensorflow.

An experiment on the performance of homemade Q-learning AIs in Agar.io depending on their state representation and available actions

The LaTeX and Python code for generating the paper, experiments' results and visualizations reported in each paper is available (whenever possible) in the paper's directory

This project uses reinforcement learning on stock market and agent tries to learn trading. The goal is to check if the agent can learn to read tape. The project is dedicated to hero in life great Jesse Livermore.

Industrial knn-based anomaly detection for images. Visit streamlit link to check out the demo.

This repo is to be freely used by ML devs to check the GAN performances without coding from scratch.

Official code for Next Check-ins Prediction via History and Friendship on Location-Based Social Networks (MDM 2018)

N-Person-Check-Checker-Splitter - A calculator app use to divide checks