FLAP: FAST, LEXICOGRAPHIC AGILE PERCEPTION INTEGRATES DECISION AND CONTROL IN A SPIKE-RESOLVED, SENSORIMOTOR PROGRAM

Abstract

Despite small neural architectures, insects make rapid, coherent, and diverse behavioral decisions in complex environments, integrating task critical information from multiple sensor modalities. In contrast, engineered systems typically rely on rich but separate modalities of sensory feedback predetermined chosen for selected tasks. This a priori design enables impressive capabilities such as automatic target recognition, tracking, and pursuit-evasion. However, the latest methods in computer vision, deep learning, and adaptive control are poorly integrated and lack animals’ abilities to parse complex and often conflicting information for multiple behaviors. The PI understands the neural basis of specialized circuits for gyroscopic stabilization, take-off, speed regulation, and prey capture. However, there is a largely untapped opportunity to understand systems-level integration of neural perception, which underlies how insects achieve both flexibility and decisiveness.

Document Details

Document Type
DoD Grant Award
Publication Date
Apr 20, 2023
Source ID
FA95502210315

Entities

People

  • Simon Sponberg

Organizations

  • Air Force Office of Scientific Research
  • Georgia Tech Research Corporation
  • United States Air Force

Tags

Readers

  • Neural Network Machine Learning.
  • Robotics and Automation.
  • Systems Analysis and Design

Technology Areas

  • AI & ML
  • AI & ML - Autonomous Systems
  • AI & ML - Neural Networks