Related papers: Flexible Bit-Truncation Memory for Approximate App…

Edge learning facilitates ubiquitous intelligence by enabling model training and adaptation directly on data-generating devices, thereby mitigating privacy risks and communication latency. However, the high computational and energy overhead…

Approximate computing (AC) leverages the inherent error resilience and is used in many big-data applications from various domains such as multimedia, computer vision, signal processing, and machine learning to improve systems performance…

Deploying large-scale transformer models on edge devices presents significant challenges due to strict constraints on memory, compute, and latency. In this work, we propose a lightweight yet effective multi-stage optimization pipeline…

In modern low-power embedded platforms, floating-point (FP) operations emerge as a major contributor to the energy consumption of compute-intensive applications with large dynamic range. Experimental evidence shows that 50% of the energy…

On-device learning enables edge devices to continually adapt the AI models to new data, which requires a small memory footprint to fit the tight memory constraint of edge devices. Existing work solves this problem by reducing the number of…

Memory-centric computing aims to enable computation capability in and near all places where data is generated and stored. As such, it can greatly reduce the large negative performance and energy impact of data access and data movement, by…

Deep neural networks for real-time video matting suffer significant computational limitations on edge devices, hindering their adoption in widespread applications such as online conferences and short-form video production. Binarization…

We propose a novel approach to reduce memory consumption of the backpropagation through time (BPTT) algorithm when training recurrent neural networks (RNNs). Our approach uses dynamic programming to balance a trade-off between caching of…

Printed electronics have gained significant traction in recent years, presenting a viable path to integrating computing into everyday items, from disposable products to low-cost healthcare. However, the adoption of computing in these…

Neural networks commonly execute on hardware accelerators such as NPUs and GPUs for their size and computation overhead. These accelerators are costly and it is hard to scale their resources to handle real-time workload fluctuations. We…

The ever-growing computational demands of increasingly complex machine learning models frequently necessitate the use of powerful cloud-based infrastructure for their training. Binary neural networks are known to be promising candidates for…

Given the recent success of Deep Learning applied to a variety of single tasks, it is natural to consider more human-realistic settings. Perhaps the most difficult of these settings is that of continual lifelong learning, where the model…

Fine-tuning pretrained large models to downstream tasks is an important problem, which however suffers from huge memory overhead due to large-scale parameters. This work strives to reduce memory overhead in fine-tuning from perspectives of…

The resource requirements of deep neural networks (DNNs) pose significant challenges to their deployment on edge devices. Common approaches to address this issue are pruning and mixed-precision quantization, which lead to latency and memory…

In this thesis I explore challenging discrete energy minimization problems that arise mainly in the context of computer vision tasks. This work motivates the use of such "hard-to-optimize" non-submodular functionals, and proposes methods…

Deeply embedded systems often have the tightest constraints on energy consumption, requiring that they consume tiny amounts of current and run on batteries for years. However, they typically execute code directly from flash, instead of the…

This paper investigates the usage of FPGA devices for energy-efficient exact kNN search in high-dimension latent spaces. This work intercepts a relevant trend that tries to support the increasing popularity of learned representations based…

The design of approximate adders has been widely researched to advance energy-efficient hardware for computation-intensive multimedia applications, such as image, audio, or video processing. The design of approximate adders has been widely…

Transformer-based models excel in speech recognition. Existing efforts to optimize Transformer inference, typically for long-context applications, center on simplifying attention score calculations. However, streaming speech recognition…

Spin Transfer Torque Random Access Memory (STT-RAM) has garnered interest due to its various characteristics such as non-volatility, low leakage power, high density. Its magnetic properties have a vital role in STT switching operations…