TENN & TENN fast are faster and more accurate than:
True North CNN, and
Loihi Slayer
on the IBM 128 DVS Data Set for Hand Gestures identification.
In fact both TENN and TENN fast produce State of the Art Performance and TENN achieves accuracy of 100 percent.
100 percent accuracy is difficult to beat just in case you were wondering.
The nearest competitor network TA-SNN falls short by 1.4% and uses three times the amount of processing time in trying to reach 100% accuracy.
Using TENN fast to recognise gestures gives a reaction time of 105 milliseconds which is about half the average reaction time of a human trained to perform a task.
Just in case you missed it the competitor networks ranged in reaction times from 155 milliseconds to 1500 milliseconds.
My opinion only DYOR
Fact Finder
TENNS, page-32
- There are more pages in this discussion • 7 more messages in this thread...
You’re viewing a single post only. To view the entire thread just sign in or Join Now (FREE)
Add BRN (ASX) to my watchlist
 (20min delay)
|
|||||
|
Last
22.5¢ |
Change
0.005(2.27%) |
Mkt cap ! $417.5M | |||
| Open | High | Low | Value | Volume |
| 22.0¢ | 23.5¢ | 22.0¢ | $975.2K | 4.296M |
Buyers (Bids)
| No. | Vol. | Price($) |
|---|---|---|
| 6 | 173129 | 22.5¢ |
Sellers (Offers)
| Price($) | Vol. | No. |
|---|---|---|
| 23.0¢ | 643724 | 13 |
View Market Depth
| Last trade - 16.10pm 07/06/2024 (20 minute delay) ? |
|
|||||
|
Last
22.8¢ |
Change
0.005 ( 2.80 %) |
||||
| Open | High | Low | Volume | ||
| 22.5¢ | 23.5¢ | 22.3¢ | 3655410 | ||
| Last updated 15.59pm 07/06/2024 ? | |||||
| BRN (ASX) Chart |