Our company is a software development company that sells an AI skeleton detection engine called VisionPose, that can detect human skeletons using AI. Since its release, this series has been embraced by over 400 companies.
In this blog, we will explain about VisionPose Nano, which is one of the series of Visionpose and how it differs from the other VisionPose series.
Kindly note that the VisionPose Nano series is exclusively available in Japan.
Introducing VisionPose
VisionPose is an AI skeleton detection engine developed by our company that can detect the depth information of multiple people using only a web camera.
It perfoms inference and analysis using deep learning (AI), allowing analysis not only from real-time footage but also videos and still images.
Features of VisionPose Nano
Low Initial Investment
VisionPose Nano is an AI skeleton detection platform equipped with our proprietary AI skeleton detection engine, VisionPose, integrated into the extremely compact hardware Jetson Nano from NVIDIA, designed for AI deep learning.
In addition, two types of skeleton detection applications are included as standard, allowing immediate use upon delivery.
As a result, we have achieved a significant reduction in size and price, making it the most affordable AI skeleton detection platform available.
The entire package, including both hardware and software, is referred to as VisionPose Nano.
No setup and installation required
There is no troublesome setup or installation as the skeleton detection application is pre-installed when the product is shipped.
30 key points! (Easily export skeletal information in .csv format)
VisionPose Nano can detect skeletal information for a total of 30 keypoints, including joints in various parts of the body (25 keypoints) and facial features (5 keypoints), all in 2D coordinates.
Additionally, the acquired data can be exported in .csv format for further analysis and processing.
Any demerits despite its convenience?
Offering deep learning AI based skeletal detection in affordable price which also includes a PC, has led to a growing demand for easy experimentation. However, there is one thing to note: VisionPose Nano is designed for research, development, and evaluation purposes, and therefore cannot be used for commercial purposes.
VisionPose Nano was created to address the common concern of needing high-spec PCs for implementing AI skeletal detection engines, which can pose challenges for research and development.
Therefore, we recommend trying skeletal detection with VisionPose Nano first. If everything works well, you may then consider exploring advanced development engines like VisionPose Standard or Single3D.
Additionally, you can also try a basic analysis with VisionPose for still images for free on our website. So, it might be a worth giving it a try there as well.
Please click the link below.
Free VisionPose AI Pose Estimation Demo
▼Click here for more information about VisionPose
Utilization in various scenes
Specific use cases for VisionPose Nano include the following:
As an AI educational material
As educational material for students, providing accessible AI skeletal detection technology.
As motion data materials
For creating works utilizing motion.
For Research purposes
Utilizing CSV data for analysis.
Specific functions
VisionPose Nano mainly has two functions installed.
VP Analyzer
Detect skeletal information from video and still images.
Skeletal information can be detected in 2D coordinates from your existing videos and still images.
The detected skeletal information can be visualized and output.
Image file types: [Input] .jpg, .png, .bmp –> [OUTPUT] .jpg, .png
Video file types: [INPUT] .avi, .mp4, .wmv, .mov –> [OUTPUT] .avi
BodyAndColor
Visualize skeleton information from camera footage.
Skeletal information can be visualized in real time and displayed on video.
※ Real-time performance is approximately 4 FPS.
Furthermore, since March, 2020, we developed High-Speed Edition BodyAndColor (β). Through lightweight and acceleration measures, processing capabilities has been increased by over six times compared to the previous version. (The conventional BodyAndColor is also included)
It allows visualization of skeletal information obtained from the camera in 2D coordinates (22 keypoints) and displaying them on the video. In addition, the detected skeletal data can also be outputted in .csv format.
※ Real-time performance is 25FPS.
Package Details
Offerings of this product
[Hardware]
- Jetson Nano unit
- AC Adaptar
- Fan
- 128GB microSD card
- Web camera
[Software]
- VisionPose Engine
- VP Analyzer
- BodyAndColor
- High-Speed Edition BodyAndColor (β)
- Fan and SD card will be delivered with the fan and SD card set into the unit.
- Monitor, keyboard, mouse, and HDMI cable will not be included.
- Camera will only work with the included Jetson Nano that you purchased.
- OS is Linux (Ubuntu).
- This product is intended for research and development and VisionPose evaluation purposes only and is not for commercial use.
Preparing VisionPose Nano for shipment
VisionPose Nano is very lightweight, but it’s also durable due to being covered with sturdy acrylic panels. It takes about less than two hours from assembly to setup completion to prepare for shipment. However, all of these tasks are completed before shipping, so you can start using it right away once you power it on, even if you’re a beginner.
Activating VisionPose Nano
Screen layout when VisionPose Nano is powered on.
After extracting images from USB or other sources, and simply drag and drop the images or videos you want to analyze for skeletal detection onto VP Analyzer and VisionPose Nano will detect the bones for you like in the image below!
When you attach a webcam to VisionPose Nano and click on BodyAndColor, it will detect the bones of the person captured by the camera.
Once you click, the detection happens instantly.
There’s absolutely no unnecessary setup required! All you need at the bare minimum is a mouse, keyboard, monitor, and HDMI cable, making it usable anywhere.
Lastly
Previously, utilizing skeleton detection systems based on deep learning (AI), including VisionPose, often required high-spec PCs, making it tough for many businesses and research instituitions at universities with limited budgets to try them out.
However, the VisionPose Nano has resolved all of these drawbacks.
Unlike other VisionPose series, it cannot be used for commercial purposes, but it serves as a valuable resource for AI learners and researchers who previously found AI skeletal detection technology too expensive and complex to adopt.
We at NextSystem will continue to advance development to contribute to the advancement of Japan’s cutting-edge technology.
You can also find us on Twitter (X), Facebook, Website, Youtube, and Linkedin.
