Advances in Machine Learning & Artificial Intelligence(AMLAI)

Smack Technologies, USA

Publications

• Review Article   
  Pulse Modulated Neural Nets with Adaptive Temporal Windows for Enhanced Video Perception
  Author(s): Greg Passmore*
  
  A key challenge in modern artificial intelligence is the limited awareness of temporal dynamics. This is especially evident in AI's struggles with video animation and interpreting sequential frames or time-dependent data. This paper presents a new method for modulating neuron spikes to synchronize with incoming frames, addressing the temporal limitations of voltage AI models. This approach involves a specialized class of neuron, similar to Wolfgang Maass's liquid state machine concept. However, our model differs significantly in its synchronization strategy. Neurons are synchronized with frame import, incorporating a flexible time window that enables spike modulation ranging from 1 to 1024 spikes per frame. This adaptability is crucial for accurately processing and interpreting time-based data. The main focus of this paper is to explore the operational prin.. Read More»
  

  Abstract HTML PDF