Shuji Ono

@shuono123
0 Followers
15 Following
4 Posts

[New Paper] HitoMi-Cam: A Shape-Agnostic Person Detection Method Using the Spectral Characteristics of Clothing

This method complements CNNs by detecting "clothing" (not shape), ideal for SAR. Runs at 23.2 fps on Raspberry Pi 5.

đź“„ Paper: https://doi.org/10.3390/jimaging11110399
🎥 Demo: https://youtu.be/612K0Kn-cKA
#HitoMiCam #ComputerVision #EdgeAI #SAR #ShapeAgnostic #SpectralImaging

HitoMi-Cam's new station demo! "Finally out in the city!" 🚀
Groundbreaking: We detect humans by clothing's spectrum (not shape!), targeting real-time on Raspberry Pi 5. An unprecedented approach!
To be presented at #SSII2025.
🎥 Video: https://youtu.be/TiI6vtEp6M8?si=DVhajczpP4rYGN3k
đź“„ Paper: https://doi.org/10.3390/jimaging11040093
#HitoMiCam #SpectralDetection #HumanDetection #RPi5 #EdgeAI #SSII2025
Real-Time Human Detection by Clothing Spectrum on Raspberry Pi 5: HitoMi-Cam Station Demo

YouTube
[New Video] Real-time Human Detection with 4-Band Raspberry Pi Camera
Check out our new demo video showing real-time human detection using a 4-Band Raspberry Pi Camera! The video displays the camera's view along with inference results highlighting detected humans.
🎥 Demo Video: https://youtu.be/oafrmEUXJYA?si=mPRbKomyCWjVvOAb
đź“„ Paper: https://doi.org/10.3390/jimaging11040093
#Multispectral #ComputerVision #EdgeAI #RaspberryPi #HumanDetection
Real-time Human Detection with 4-Band Raspberry Pi Camera

YouTube
[New Paper] A Color‑Based Multispectral Imaging Approach for a Human‑Detection Camera
“HitoMi‑Cam” uses a 4‑band snapshot filter + Raspberry Pi to detect humans in real‑time (15 fps).
đź“„ Paper: https://doi.org/10.3390/jimaging11040093
🎥 Demo: https://youtu.be/VE2gudrNOCs?si=HBEnEruAqIxxN1Dt
#Multispectral #ComputerVision #EdgeAI