Related Experiment Videos
Spiking Neural Networks with Continual Learning for Steering Angle Regression: A Sustainable AI Perspective.
Fernando S Martínez1, Sergio Costa1, Raúl Parada2
1e-Health Center, Universitat Oberta de Catalunya (UOC), 08018 Barcelona, Spain.
Sensors (Basel, Switzerland)
|May 13, 2026
Summary
Spiking Neural Networks (SNNs) with Continual Learning (CL) effectively perform steering angle regression in autonomous driving simulations. This approach enhances energy efficiency and reduces CO2 emissions by optimizing data usage.
Area of Science:
- Artificial Intelligence
- Sustainable Computing
- Autonomous Systems
Background:
- Deep learning models face increasing energy demands, challenging sustainability.
- Spiking Neural Networks (SNNs) offer computational and energy efficiency, inspired by biological systems.
- Continual Learning (CL) is crucial for sequential tasks to prevent catastrophic forgetting.
Purpose of the Study:
- Apply SNNs and CL to steering angle regression in autonomous driving simulations.
- Focus on energy efficiency and reducing CO2 emissions.
- Assess SNNs' ability to maintain accuracy while optimizing data usage.
Main Methods:
- Adapted PilotNet architecture for SNNs.
- Utilized datasets from the Udacity driving simulator.
- Implemented CL techniques like Elastic Weight Consolidation and replay memory.
- Evaluated models in incremental learning scenarios.
Main Results:
- SNNs with replay memory retained prior knowledge with minimal energy increase.
- Compared SNNs with CL against baseline models using MSE, efficiency, and CO2 emissions.
- Demonstrated the viability of SNNs for sustainable AI applications.
Conclusions:
- SNNs integrated with CL present a sustainable alternative for AI.
- Replay memory is effective in mitigating catastrophic forgetting in SNNs.
- Future work should explore hardware implementations and real-world testing.
Related Concept Videos
01:37Neural Regulation
Digestion begins with a cephalic phase that prepares the digestive system to receive food. When our brain processes visual or olfactory information about food, it triggers impulses in the cranial nerves innervating the salivary glands and stomach to prepare for food.
01:12Observational Learning
Albert Bandura's observational learning, also known as imitation or modeling, occurs when a person observes and imitates another's behavior. It is a quicker process than operant conditioning. A well-known example is the Bobo doll study, where children who saw an adult acting aggressively towards the doll were more likely to act aggressively when left alone, compared to those who observed a nonaggressive adult. Many psychologists view observational learning as a form of latent learning because...