Projects
A selective overview of some research projects of mine.
2024
- MELTing Point: Mobile Evaluation of Language Transformers
- Recurrent Early Exits for Federated Learning with Heterogeneous Clients
- Maestro: Uncovering Low-Rank Structures via Trainable Decomposition
2023
- Fluid Batching: Exit-Aware Preemptive Serving of Early-Exit Neural Networks on Edge NPUs
- FedorAS: Federated Architecture Search under system heterogeneity
2022
- MESS: Multi-Exit Semantic Segmentation Networks
- DynO: Dynamic Onloading of Deep Neural Networks from Cloud to Device
2021
- FjORD: Fair and Accurate Federated Learning under heterogeneous targets with Ordered Dropout
- Smart at what cost? Characterising Mobile Deep Neural Networks in the wild
- It's always personal: Using Early Exits for Efficient On-Device CNN Personalisation
