Scenario 8
License
If you want to use the dataset or scripts on this page, please use the link below to generate the final list of papers that need to be cited.
Overview
Scenario 8 emulates a Vehicle-to-Infrastructure (V2I) mmWave communication setup. The adopted testbed comprises of three units. Unit 1 primarily consists of a stationary base station equipped with an RGB camera, a LiDAR sensor, and a mmWave phased array. The stationary unit adopts a 16-element 60GHz-band phased array and it receives the transmitted signal using an over-sampled codebook of 64 pre-defined beams. The second unit (Unit 2) is a mobile vehicle unit equipped with a mmWave transmitter and GPS receiver. The transmitter consists of a quasi-omni antenna constantly transmitting (omnidirectional) at the 60 GHz band. The scanning range of the LiDAR is 16 meters and the motor spin frequency is 10Hz. Each data sample has multiple modalities including an RGB image and a LiDAR 360-degree point cloud, both collected by Unit 1. Please refer to the detailed description of the testbed presented here.
McAllister Ave: The testbed is deployed during the daytime at McAllister avenue, the same street as that of Scenarios 1, 2, and 5, but a little south to other locations. It is a two-way street with 2 lanes, a width of 10.6 meters, and a vehicle speed limit of 25mph (40.6 km per hour. The stationary unit is placed close to the entrance of a vehicle parking lot. This location has vehicle traffic and also footfall and cycling traffic. Vehicles can be seen driving through the street or driving into or out of the parking structure. All of these features make this scenario diverse from visual and wireless perspectives alike.
Collected Data
Overview
Number of Data Collection Units: 2 (using DeepSense Testbed #1)
Number of Data Samples: 4043
Data Modalities: RGB images, 64-dimensional received power vector, GPS locations
Average Data Capture Rate: 7.79 FPS
Sensors at Unit 1: (Stationary Receiver)
- Wireless Sensor [Phased Array]: A 16-element antenna array operating in the 60 GHz frequency band and receives the transmitted signal using an over-sampled codebook of 64 pre-defined beams
- Visual Sensor [Camera]: The main visual perception element in the testbed is an RGB-D camera. The camera is used to capture RGB images of 960×540 resolution at a base frame rate of 30 frames per second (fps)
- Position Sensor [GPS Receiver]: A GPS-RTK receiver for capturing accurate real-time locations for the stationary unit.
- LiDAR Sensor [2D laser scanner]: This is a co-located system and it provides the range and angle data of objects within its range of view. This system covers a maximum range of 40 meters around the master component and it operates at a maximum scan rate of 10 Hz.
Sensors at Unit 2: (Mobile Transmitter)
- Position Sensor [GPS Receiver]: A GPS-RTK receiver is installed on the top of the mobile unit and is used to capture accurate real-time locations at 10 frames per second (fps). The collected data comprises the Latitude and Longitude information in addition to other important data.
| Testbed | 1 |
|---|---|
| Instances | 4043 |
| Number of Units | 2 |
| Data Modalities | RGB images, 64-dimensional received power vector, GPS locations, LiDAR point cloud |
| Unit1 | |
| Type | Stationary |
| Hardware Elements | RGB camera, mmWave phased array receiver, GPS receiver, LiDAR |
| Data Modalities | RGB images, 64-dimensional received power vector, GPS locations, LiDAR point cloud |
| Unit2 | |
| Type | Mobile |
| Hardware Elements | mmWave omni-directional transmitter, GPS receiver |
| Data Modalities | GPS locations |
Data Visualization
Download
Please login to download the DeepSense datasets
How to Access Scenario 8 Data?
Step 1. Download Scenario 8 Data
Step 2. Extract the scenario8.zip file
Scenario 8 folder consists of three sub-folders:
- unit1: Includes the data captured by unit 1
- unit2: Includes the data captured by unit 2
- resources: Includes the scenario-specific annotated dataset, data labels and other additional information. For more details, refer the resources section below.
Scenario 8 folder also includes the “scenario 8.csv” file with the paths to all the collected data. For each coherent time, we provide the corresponding visual, wireless and GPS data.
Resources
What are the Additional Resources?
Resources consist of the following information:
- visual data annotations: For the visual data, we provide the coordinates of the 2D bounding box and attributes for each frame
- data labels: The labels comprises of the ground-truth beam indices computed from the mmWave received power vectors, the direction of travel (unit2), and the sequence index
- additional information: Includes the scenario-specific additional data. Details of the information is provided below
Visual Data Annotations
After performing the post-processing steps presented here, we generate the annotations for the visual data. Using state-of-the-art machine learning algorithms and multiple validation steps, we achieve highly accurate annotations. In this particular scenario, we provide the coordinates of the 2D bounding box and attributes for each frame. We, also, provide the ground-truth labels for 2 object classes, “Tx”, and “Distractor”. The “Tx” refers to the transmitting vehicle in the scene and “Distractor” for any other objects, such as human, other vehicles, etc. We follow the YOLO format for the bounding-box information. In the YOLO format, each bounding box is described by the center coordinates of the box and its width and height. Each number is scaled by the dimensions of the image; therefore, they all range between 0 and 1. Instead of category names, we provide the corresponding integer categories. We follow the following assignment: (i) “Tx” as “0” , and (ii) “Distractor” as “1”.
Data Labels
The labels comprises of the ground-truth beam indices computed from the mmWave received power vectors, the direction of travel (unit2), and the sequence index.
- Ground-Truth Beam: The phased array of unit 1 utilizes an over-sampled beamforming codebook of N = 64 vectors, which are designed to cover the field of view. It captures the received power by applying the beamforming codebook elements as a combiner. For each received power vector of dimension [64 x 1], the index with the maximum received power value is selected as the optimal beam index. This data is provided in the column 10 [‘unit1_beam_index’] of the scenario8.csv
- Sequence Index: During the data collection process, the mobile transmitter (unit2) travelled multiple times in front of the base station (unit1). For each run, the testbed collects multiple data samples. All the data samples with the same sequence index belongs to the same run
- Direction of Travel: For this scenario, during the data collection process, the mobile unit generally moves either from left-to-right or right-to-left of the base station. Here “0” represents the left-to-right movement of the transmitter and “1” represents the “right-to-left” movement. The movement is measured is from the point of view of the base station. This is provided in the ‘unit2_direction’ column of the csv
Additional Information
We, further, provide additional information for each sample present in the scenario dataset. The details are provided in the columns 11 – 13 of the scenario8.csv. The contents of the additional data is listed below:
- index: It represents the sample number
- time_stamp[UTC]: This represents the time of data capture in “hr-mins-secs-ms” format
An example table comprising of the data labels and the additional information is shown below.
| index | unit1_beam_index | seq_index | time_stamp[UTC] | unit2_direction |
|---|---|---|---|---|
| 1 | 15 | 1 | ['22-11-14-0'] | 0 |
| 2 | 15 | 1 | ['22-11-14-250'] | 0 |
| 3 | 2 | 1 | ['22-11-14-500'] | 0 |
| 4 | 2 | 1 | ['22-11-14-750'] | 0 |
| 5 | 2 | 1 | ['22-11-15-0'] | 0 |
| 6 | 2 | 1 | ['22-11-15-111'] | 0 |
| 7 | 2 | 1 | ['22-11-15-222'] | 0 |
| 8 | 3 | 1 | ['22-11-15-333'] | 0 |
| 9 | 4 | 1 | ['22-11-15-444'] | 0 |
| 10 | 5 | 1 | ['22-11-15-555'] | 0 |
