88 citations · 272 across the 16 of their papers we have counts for
17 papers
Empowering On-Device Model Adaptation with an Edge AI Inference Accelerator
Mateusz Piechocki, Alessandro Capotondi, Marek Kraft
On-device model adaptation is essential to enable lifelong personalization on resource-constrained hardware, but compute, power, and memory limitations of such devices make end-to-…
Diffusion Is Your Friend in Show, Suggest and Tell
Jia Cheng Hu, Roberto Cavicchioli, Alessandro Capotondi
Diffusion Denoising models demonstrated impressive results across generative Computer Vision tasks, but they still fail to outperform standard autoregressive solutions in the discr…
Shifted Window Fourier Transform And Retention For Image Captioning
Jia Cheng Hu, Roberto Cavicchioli, Alessandro Capotondi
Image Captioning is an important Language and Vision task that finds application in a variety of contexts, ranging from healthcare to autonomous vehicles. As many real-world applic…
Bidirectional Awareness Induction in Autoregressive Seq2Seq Models
Jia Cheng Hu, Roberto Cavicchioli, Alessandro Capotondi
Autoregressive Sequence-To-Sequence models are the foundation of many Deep Learning achievements in major research fields such as Vision and Natural Language Processing. Despite th…
Heterogeneous Encoders Scaling In The Transformer For Neural Machine Translation
Jia Cheng Hu, Roberto Cavicchioli, Giulia Berardinelli +1
Although the Transformer is currently the best-performing architecture in the homogeneous configuration (self-attention only) in Neural Machine Translation, many State-of-the-Art m…
The Importance of Worst-Case Memory Contention Analysis for Heterogeneous SoCs
Lorenzo Carletti, Gianluca Brilli, Alessandro Capotondi +2
Memory interference may heavily inflate task execution times in Heterogeneous Systems-on-Chips (HeSoCs). Knowing worst-case interference is consequently fundamental for supporting…
A request for clarity over the End of Sequence token in the Self-Critical Sequence Training
Jia Cheng Hu, Roberto Cavicchioli, Alessandro Capotondi
The Image Captioning research field is currently compromised by the lack of transparency and awareness over the End-of-Sequence token (<Eos>) in the Self-Critical Sequence Training…
HULK-V: a Heterogeneous Ultra-low-power Linux capable RISC-V SoC
Luca Valente, Yvan Tortorella, Mattia Sinigaglia +4
IoT applications span a wide range in performance and memory footprint, under tight cost and power constraints. High-end applications rely on power-hungry Systems-on-Chip (SoCs) fe…
Exploiting Multiple Sequence Lengths in Fast End to End Training for Image Captioning
Jia Cheng Hu, Roberto Cavicchioli, Alessandro Capotondi
We introduce a method called the Expansion mechanism that processes the input unconstrained by the number of elements in the sequence. By doing so, the model can learn more effecti…
Exploring the sequence length bottleneck in the Transformer for Image Captioning
Jia Cheng Hu, Roberto Cavicchioli, Alessandro Capotondi
Most recent state of the art architectures rely on combinations and variations of three approaches: convolutional, recurrent and self-attentive methods. Our work attempts in laying…
A TinyML Platform for On-Device Continual Learning with Quantized Latent Replays
Leonardo Ravaglia, Manuele Rusci, Davide Nadalini +3
In the last few years, research and development on Deep Learning models and techniques for ultra-low-power devices in a word, TinyML has mainly focused on a train-then-deploy assum…
Leveraging Automated Mixed-Low-Precision Quantization for tiny edge microcontrollers
Manuele Rusci, Marco Fariselli, Alessandro Capotondi +1
The severe on-chip memory limitations are currently preventing the deployment of the most accurate Deep Neural Network (DNN) models on tiny MicroController Units (MCUs), even if le…
Memory-Latency-Accuracy Trade-offs for Continual Learning on a RISC-V Extreme-Edge Node
Leonardo Ravaglia, Manuele Rusci, Alessandro Capotondi +5
AI-powered edge devices currently lack the ability to adapt their embedded inference models to the ever-changing environment. To tackle this issue, Continual Learning (CL) strategi…
Robustifying the Deployment of tinyML Models for Autonomous mini-vehicles
Miguel de Prado, Manuele Rusci, Romain Donze +4
Standard-size autonomous navigation vehicles have rapidly improved thanks to the breakthroughs of deep learning. However, scaling autonomous driving to low-power systems deployed o…
Memory-Driven Mixed Low Precision Quantization For Enabling Deep Network Inference On Microcontrollers
Manuele Rusci, Alessandro Capotondi, Luca Benini
This paper presents a novel end-to-end methodology for enabling the deployment of low-error deep networks on microcontrollers. To fit the memory and computational limitations of re…
HERO: Heterogeneous Embedded Research Platform for Exploring RISC-V Manycore Accelerators on FPGA
Andreas Kurth, Pirmin Vogel, Alessandro Capotondi +2
Heterogeneous embedded systems on chip (HESoCs) co-integrate a standard host processor with programmable manycore accelerators (PMCAs) to combine general-purpose computing with dom…
NEURAghe: Exploiting CPU-FPGA Synergies for Efficient and Flexible CNN Inference Acceleration on Zynq SoCs
Paolo Meloni, Alessandro Capotondi, Gianfranco Deriu +5
Deep convolutional neural networks (CNNs) obtain outstanding results in tasks that require human-level understanding of data, like image or speech recognition. However, their compu…