A5004225952- Advanced Memory and Neural Computing
- Neural dynamics and brain function
- Neurobiology and Insect Physiology Research
- Cell Image Analysis Techniques
- Insect and Arachnid Ecology and Behavior
- Neural Networks and Reservoir Computing
- Plant and animal studies
- Ferroelectric and Negative Capacitance Devices
- Photoreceptor and optogenetics research
Johns Hopkins University Applied Physics Laboratory
2022-2024
The learning center in the insect, mushroom body (MB) with its predominant population of Kenyon Cells (KCs), is a widely studied model system to investigate neural processing principles, both experimentally and theoretically. While many computational models MB have been studied, role recurrent connectivity between KCs remains inadequately understood. Dynamical point attractors are candidate theoretical framework where connections network can enable discrete set stable activation patterns....
Continuous state estimation is a fundamental problem addressed by neural systems which underlies complex capabilities such as navigation. While studies in heading direction of the fruit fly D. melanogaster have uncovered how this computation can be performed ring attractor architecture, it unclear additional synapse-level architectural details contribute to functional performance estimation. In work, we find consistent, repeated, motif connectivity data captures distributions connections...
Despite the progress in deep learning networks, efficient at edge (enabling adaptable, low-complexity machine solutions) remains a critical need for defense and commercial applications. We envision pipeline to utilize large neuroimaging datasets, including maps of brain which capture neuron synapse connectivity, improve approaches. have pursued different approaches within this structure. First, as demonstration data-driven discovery, team has developed technique discovery repeated...
Despite the progress in deep learning networks, efficient at edge (enabling adaptable, low-complexity machine solutions) remains a critical need for defense and commercial applications. We have pursued multiple neuroscience-inspired AI efforts which may overcome these data inefficiencies, power lack of generalization. envision pipeline to utilize large neuroimaging datasets, including maps brain capture neuron synapse connectivity, improve approaches. different approaches within this...
Continual learning without catastrophic forgetting of previous experiences is an open general challenge for artificial neural networks, but especially under-explored networks suitable to implement on neuromorphic platforms. An algorithmic understanding how continual occurs in biological can inform solutions platforms whose biomimetic computing architectures lend themselves more biofidelic algorithms. In this work, we derive approach generative-replay-based with a three-factor local rule...