Xiuying Wei, Yunchen Zhang, Yuhang Li, Xiangguo Zhang, Ruihao Gong, Jinyang Guo, Xianglong Liu. Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing. 2023.

In recent years, artificial intelligence (AI) has achieved significant progress and remarkable advancements across various disciplines, including biology, computer science, and industry. However, the increasing complexity of AI network structures and the vast number of associated parameters impose substantial computational and storage demands, severely limiting the practical deployment of these models on resource-constrained edge devices. Although edge intelligence methods have been proposed to alleviate the computational and storage burdens, they still face multiple persistent challenges, such as large-scale model deployment, poor interpretability, privacy and security vulnerabilities, and energy efficiency constraints. This article systematically reviews the current advancements in edge intelligence technologies, highlights key enabling techniques including model sparsity, quantization, knowledge distillation, neural architecture search, and federated learning, and explores their applications in industrial, automotive, healthcare, and consumer domains. Furthermore, this paper presents a comparative analysis of these techniques, summarizes major trade-offs, and proposes decision frameworks to guide deployment strategies under different scenarios. Finally, it discusses future research directions to address the remaining technical bottlenecks and promote the practical and sustainable development of edge intelligence. Standing at the threshold of an exciting new era, we believe edge intelligence will play an increasingly critical role in transforming industries and enabling ubiquitous intelligent services.

Voting-based methods (e.g., VoteNet) have achieved promising results for 3D object detection. However, the simple voting operation in VoteNet may lead to less accurate voting results that are far away from the true object centers. In this work, we propose a simple but effective 3D object detection method called Transformer3D-Det (T3D), in which we additionally introduce a transformer based vote refinement module to refine the voting results of VoteNet and can thus significantly improve the 3D object detection performance. Specifically, our T3D framework consists of three modules: a vote generation module, a vote refinement module, and a bounding box generation module. Given an input point cloud, we first utilize the vote generation module to generate multiple coarse vote clusters. Then, the clustered coarse votes will be refined by using our transformer based vote refinement module to produce more accurate and meaningful votes. Finally, the bounding box generation module takes the refined vote clusters as the input and generates the final detection result for the input point cloud. To alleviate the impact of inaccurate votes, we also propose a new non-vote loss function to train our T3D. As a result, our T3D framework can achieve better 3D object detection performance. Comprehensive experiments on two benchmark datasets ScanNetV2 and SUN RGB-D demonstrate the effectiveness of our T3D framework for 3D object detection.

In this work, we propose a simple but effective channel pruning framework called Progressive Channel Pruning (PCP) to accelerate Convolutional Neural Networks (CNNs). In contrast to the existing channel pruning methods that prune channels only once per layer in a layer-by-layer fashion, our new progressive framework iteratively prunes a small number of channels from several selected layers, which consists of a threestep attempting-selecting-pruning pipeline in each iteration. In the attempting step, we attempt to prune a pre-defined number of channels from one layer by using any existing channel pruning methods and estimate the accuracy drop for this layer based on the labelled samples in the validation set. In the selecting step, based on the estimated accuracy drops for all layers, we propose a greedy strategy to automatically select a set of layers that will lead to less overall accuracy drop after pruning these layers. In the pruning step, we prune a small number of channels from these selected layers. We further extend our PCP framework to prune channels for the deep transfer learning methods like Domain Adversarial Neural Network (DANN), in which we effectively reduce the data distribution mismatch in the channel pruning process by using both labelled samples from the source domain and pseudo-labelled samples from the target domain. Our comprehensive experiments on two benchmark datasets demonstrate that our PCP framework outperforms the existing channel pruning approaches under both supervised learning and transfer learning settings.

In this work, we propose a new deep image compression framework that aims to learn one single network to support variable bitrate coding under various computational complexity levels. In contrast to the existing state-of-the-art learning-based image compression frameworks that only consider the rate-distortion trade-off without introducing any constraint related to the computational complexity, our Complexity and Bitrate Adaptive Network (CBANet) considers the complex rate-distortion-complexity trade-off when learning a single network to support multiple computational complexity levels and variable bitrates. Since it is a non-trivial task to solve such a rate-distortion-complexity related optimization problem, we propose a two-step approach to decouple this complex optimization task into a complexity-distortion optimization sub-task and a rate-distortion optimization sub-task, and additionally propose a new network design strategy by introducing a Complexity Adaptive Module (CAM) and a Bitrate Adaptive Module (BAM) to respectively achieve the complexity-distortion and rate-distortion trade-offs. As a general approach, our network design strategy can be readily incorporated into different deep image compression methods to achieve complexity and bitrate adaptive image compression by using a single network. Comprehensive experiments on two benchmark datasets demonstrate the effectiveness of our CBANet for deep image compression. Code is released at https://github.com/JinyangGuo/CBANet-release.