Looking a for Facial Recognition Hardware Tester!

I would love to build and demonstrate usage of such great device in the vehicle. One set up would be drowsiness detection in attempt to keep the driver safe and awake, second would be an interesting approach on security of the vehicle using facial recognition. Other, use cases could also be control of the radio based on the user facial expression, if the driver seems mad maybe play some relaxing music.

I would make it a facial lock for my bedroom

Hi Phil,

I am a researcher in Robotics. Since the past few years, I have been working on various projects that involve autonomous systems, embedded systems and computer vision. So far I have only been working with Rpi cameras and USB cameras for computer vision applications. Just came across the HVC system which could be used for such kind of applications. I am curious to integrate this into one of my ongoing projects which involves image processing using Raspberry Pi and Rpi cam. I am doubly excited to find that this HVC system has inbuilt processing capability and works with Rpi too. It will take much of the processing load from the raspberry allowing it to run other critical tasks. Let me know if a piece is available for testing.

The following are already a part of my project requirement.

• Face Detection
• Face Recognition

• Facial Pose Estimation

• Hand Detection
• Human Body Detection

Nevertheless, I will be happy to include the remaining part too. I can also include the results in my project pages or as a separate review of the product itself.

Thanks and regards,

Cris

Hi Phil,

I'm new to this community, I joined for eagle library needs and stumbled upon your post.

This omron product seems to be amazing and as a Fab Lab and coworking manager we have plenty of projects to prototype with it.

Previously we developed a project in opencv that consisted in a dummy ( manikin ) with servos in the head that basically keep staring people passing by.

It was placed in a shop window and people had a lot of fun!

For us it would be worth to extend the functionalities with some kind of reaction based on people expressions

I've got three ideas, in order of difficulty:

• I teach in a post-graduate setting in which we would like to monitor attendance.  I've tried a number of options, such as RFID, fingerprint ID, etc.  I would be interested in trying the facial recognition as a way to track individual attendance.
• In this setting, we have a public area with public service (health) materials.  I would be interested in customizing the materials to the audience, by having the system detect age, gender, number of people, etc.
• Although this one would require IRB approval and patient consent, it would be fascinating to have this device analyze people's appearance (age, gender, expression, gaze, pose, etc) and then see if we can establish correlations with their actual age, stated gender, and medical problems.  I would be most interested to see if the results were to show an ability to detect certain issues.  For example, would this device have the sensitivity to identify depression through a combination of gaze, pose, and expression?

Given the varying levels of difficulty, I would not promise that I could complete all three.  Rather, I would work in that direction and see where it goes.

What do you think?

Id love to hear your idea and experience. We are making sure we get a few pitches for uses, We know it works with Raspberry Pi too, so that would be a preferred benefit. 

Thanks!

Sure, would love to help out.  What do you need from me to move forward?

Hey fpobdoc I have another one land on my desk today if you are interested in pitching it!

Is this still available?  I have a project but wouldn't want to write it up if the opportunity has closed.

Thanks,

Bill

I just recently got into making projects with face detection and would say the resolution on this would be better then what I am using right now.

I would use it more for home security and or for facial recognition for other applications with facial recognition with using something of a higher quality face mapping and detection ticks twitches and other things could be use for interrogation methods in real time would be cool to play with in a conversation.  Basically with a unit like this would mean projects would come to life better and be able to distinguish between items and features with more ease and less false positives which would be awesome.

To be honest given the change I would love to rip through a bunch of different projects and see how it preforms vs other cameras and devices as well as pushing the application of "quality control" in image use.  currently working on a setup with a 360 degree tracker which would be interesting to use as a poi addition to it.