Been reviewing and reading articles which all claim to use AI to identify movement patterns in animals and humans but seem to miss the key point, that CNS activity and the output seen as a motor event like walking or speech differ in temporal scales. Neuronal firing is always in milliseconds while the output or endpoint (motor event) is in seconds even when at its best.
Another key issue is that the signal recorded along axons and sites away from source are have a delay as these signals are not transmitted at the speed of light, but each axon has its own conduction velocity and it matters!!
Neurophysiologists have been examining these things for over a century now, so please do check some of the matter before jumping to conclusion, this is also for the reviewers.
Dear Samit Chakrabarty
Hand gesture recognition is one of the most sophisticated areas in which computer vision and artificial intelligence have aided not just in improving communication with deaf individuals but also in supporting gesture-based signaling systems.
Gesture recognition is a sort of touchless user interface (TUI). TUI devices, unlike touchscreen devices, are controlled without the need of touch. TUIs include voice-controlled smart speakers such as Google Home and Amazon Alexa.
The RF-Pose artificial intelligence-powered technology bounces WiFi signals through walls and off individuals on the opposite side of them, then analyzes the patterns that emerge. The technique can recreate a 2D stick figure from these blobs of reflected information.
Kind Regards
Qamar Ul Islam
Dear Qamar,
Not sure that you follow the point I make, as your arguments are absolutely fine but does not address the concerns I raise about the assumptions used to develop these toolsets and validate them.
best wishes
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Artificial Intelligence