Combining Symbolic and Neural Learning
Machine Learning
Constructing deterministic finite-state automata in recurrent neural networks
Journal of the ACM (JACM)
Motion features to enhance scene segmentation in active visual attention
Pattern Recognition Letters
Parallel algorithms development for programmable logic devices
Advances in Engineering Software
Visual surveillance by dynamic visual attention method
Pattern Recognition
Accelerating colour space conversion on reconfigurable hardware
Image and Vision Computing
Hi-index | 0.00 |
Recently, the use of the algorithmic lateral inhibition (ALI) method in motion detection has shown to be very effective. The promising results in terms of the goodness of the silhouettes detected and tracked along video sequences lead us to accept the challenge of searching for a real-time implementation of the algorithms. This paper introduces two steps towards that direction: (a) A simplification of the general ALI method is performed by formally transforming it into a finite state machine. (b) A hardware implementation of such a designed ALI module, as well as an 8×8 ALI module, has been tested on several video sequences, providing promising performance results.