- Company focuses on ultra-low-power inference on the edge
- Hardware is digital asynchronous for Speck and digital synchronous for
-
- Designing the machine learning pipeline to train our models
- Focus on engineering the tools for gradient-based optimization of SNNs, data wrangling
- Establish the use of good practices to make training and deployment faster
- Training SNNs for better performance through data augmentation
- Architectural exploration such as recurrence in the network, spike frequency adaptation
- How to save energy, fewer spikes is the goal
- Training Spiking Neural Networks Using Lessons from Deep Learning: Eshragian, Neftci, Wang, Lenz
- Adversarial Attacks on Spiking Convolutional Networks: Büchel, Lenz, Sheik, Sorbaro
- EXODUS: Stable and Efficient Training of Spiking Neural Networks: Bauer, Lenz, Haghighatshoar, Sheik
- under submission: Ultra-low-power image classification on neuromorphic hardware
- Converting an ANN to an SNN using temporal coding and a single spike per neuron
- A method to normalise ANN activation which scales to deeper layers
- CapoCaccia workshop: presented our chips, gave tutorials and demos
- AMLD: a full-day workshop with our hardware, outreach to new students
- Open Neuromorphic: Platform for neuromorphic open source code and hardware
- NeuroBench: benchmarking and metrics for neuromorphic computing
- Leading role in the algorithms vision team
- Strong engineering background
- 3 papers published in the last year, 1 under submission
- 2 patents submitted
- Involved in numerous community efforts
Made with Slides.com