An official website of the United States government
With increasingly deployed cameras and the rapid advances of Computer Vision, large-scale live video analytics becomes feasible. However, analyzing videos is compute-intensive. In addition, live video analytics needs to be performed in real time. In this paper, we design an edge server system for live video analytics. We propose to perform configuration adaptation without profiling video online. We select configurations with a prediction model based on object movement features. In addition, we reduce the latency through resource orchestration on video analytics servers. The key idea of resource orchestration is to batch inference tasks that use the same CNN model, and schedule tasks based on a priority value that estimates their impact on the total latency. We evaluate our system with two video analytic applications, road traffic monitoring and pose detection. The experimental results show that our profiling-free adaptation reduces the workload by 80% of the state-of-the-art adaptation without lowering the accuracy. The average serving latency is reduced by up to 95% comparing with the profiling-based adaptation.
more »
« less
This content will become publicly available on February 28, 2026
Pruning One More Token is Enough: Leveraging Latency-Workload Non-Linearities for Vision Transformers on the Edge
This paper investigates how to efficiently deploy vision transformers on edge devices for small workloads. Recent methods reduce the latency of transformer neural networks by removing or merging tokens, with small accuracy degradation. However, these methods are not designed with edge device deployment in mind: they do not leverage information about the latency-workload trends to improve efficiency. We address this shortcoming in our work. First, we identify factors that affect ViT latency-workload relationships. Second, we determine token pruning schedule by leveraging non-linear latency-workload relationships. Third, we demonstrate a training-free, token pruning method utilizing this schedule. We show other methods may increase latency by 2-30%, while we reduce latency by 9-26%. For similar latency (within 5.2% or 7ms) across devices we achieve 78.6%-84.5% ImageNet1K accuracy, while the state-of-the-art, Token Merging, achieves 45.8%-85.4%.
more »
« less
- Award ID(s):
- 2104709
- PAR ID:
- 10638751
- Publisher / Repository:
- The Computer Vision Foundation.
- Date Published:
- Subject(s) / Keyword(s):
- computer vision, token pruning
- Format(s):
- Medium: X
- Location:
- Tucson Arizona
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Mobile or FPGA? A Comprehensive Evaluation on Energy Efficiency and a Unified Optimization FrameworkEfficient deployment of Deep Neural Networks (DNNs) on edge devices (i.e., FPGAs and mobile platforms) is very challenging, especially under a recent witness of the increasing DNN model size and complexity. Model compression strategies, including weight quantization and pruning, are widely recognized as effective approaches to significantly reduce computation and memory intensities, and have been implemented in many DNNs on edge devices. However, most state-of-the-art works focus on ad-hoc optimizations, and there lacks a thorough study to comprehensively reveal the potentials and constraints of different edge devices when considering different compression strategies. In this paper, we qualitatively and quantitatively compare the energy efficiency of FPGA-based and mobile-based DNN executions using mobile GPU and provide a detailed analysis. Based on the observations obtained from the analysis, we propose a unified optimization framework using block-based pruning to reduce the weight storage and accelerate the inference speed on mobile devices and FPGAs, achieving high hardware performance and energy-efficiency gain while maintaining accuracy.more » « less
-
It is challenging to deploy 3D Convolutional Neural Networks (3D CNNs) on mobile devices, specifically if both real-time execution and high inference accuracy are in demand, because the increasingly large model size and complex model structure of 3D CNNs usually require tremendous computation and memory resources. Weight pruning is proposed to mitigate this challenge. However, existing pruning is either not compatible with modern parallel architectures, resulting in long inference latency or subject to significant accuracy degradation. This paper proposes an end-to-end 3D CNN acceleration framework based on pruning/compilation co-design called Mobile-3DCNN that consists of two parts: a novel, fine-grained structured pruning enhanced by a prune/Winograd adaptive selection (that is mobile-hardware-friendly and can achieve high pruning accuracy), and a set of compiler optimization and code generation techniques enabled by our pruning (to fully transform the pruning benefit to real performance gains). The evaluation demonstrates that Mobile-3DCNN outperforms state-of-the-art end-to-end DNN acceleration frameworks that support 3D CNN execution on mobile devices, Alibaba Mobile Neural Networks and Pytorch-Mobile with speedup up to 34 × with minor accuracy degradation, proving it is possible to execute high-accuracy large 3D CNNs on mobile devices in real-time (or even ultra-real-time).more » « less
-
Unstructured neural network pruning is an effective technique that can significantly reduce theoretical model size, computation demand and energy consumption of large neural networks without compromising accuracy. However, a number of fundamental questions about pruning are not answered yet. For example, do the pruned neural networks contain the same representations as the original network? Is pruning a compression or evolution process? Does pruning only work on trained neural networks? What is the role and value of the uncovered sparsity structure? In this paper, we strive to answer these questions by analyzing three unstructured pruning methods (magnitude based pruning, post-pruning re-initialization, and random sparse initialization). We conduct extensive experiments using the Singular Vector Canonical Correlation Analysis (SVCCA) tool to study and contrast layer representations of pruned and original ResNet, VGG, and ConvNet models. We have several interesting observations: 1) Pruned neural network models evolve to substantially different representations while still maintaining similar accuracy. 2) Initialized sparse models can achieve reasonably good accuracy compared to well engineered pruning methods. 3) Sparsity structures discovered by pruning models are not inherently important or useful.more » « less
-
Binary neural networks (BNNs) substitute complex arithmetic operations with simple bit-wise operations. The binarized weights and activations in BNNs can drastically reduce memory requirement and energy consumption, making it attractive for edge ML applications with limited resources. However, the severe memory capacity and energy constraints of low-power edge devices call for further reduction of BNN models beyond binarization. Weight pruning is a proven solution for reducing the size of many neural network (NN) models, but the binary nature of BNN weights make it difficult to identify insignificant weights to remove. In this paper, we present a pruning method based on latent weight with layer-level pruning sensitivity analysis which reduces the over-parameterization of BNNs, allowing for accuracy gains while drastically reducing the model size. Our method advocates for a heuristics that distinguishes weights by their latent weights, a real-valued vector used to compute the pseudogradient during backpropagation. It is tested using three different convolutional NNs on the MNIST, CIFAR-10, and Imagenette datasets with results indicating a 33%--46% reduction in operation count, with no accuracy loss, improving upon previous works in accuracy, model size, and total operation count.more » « less
