Data acquisition and processing (DAP) by uncrewed aerial vehicle (UAV) hold the key to a diverse range of practical applications. However, UAVs often face challenges in completing complex DAP tasks due to inadequate on-board resources. To ameliorate such difficulties, previous researches have explored the mobile edge computing (MEC)-assisted UAV DAP, but they do not prioritize the mobility of MEC servers, resulting in challenges when communication and computation resources are limited. To address this issue, this work proposes a novel design that incorporates a movable uncrewedground vehicle (UGV)-mounted MEC server to assist UAV DAP, which enables the UAV to thoroughly leverage edge resources. Under this design, we aim to maximize UAV DAP ability with constraints including real-time data processing, UGV obstacles avoidance and system motion. Efficient algorithms are proposed, with obstacles avoidance problem solved, from a systematic perspective. Extensive simulation and real-world experiment results demonstrate the effectiveness of our system design and algorithms.

We propose an end-to-end generative adversarial network that allows for controllable ink wash painting generation from sketches by specifying the colors via color hints. To the best of our knowledge, this is the first study for interactive Chinese ink wash painting colorization from sketches. To help our network understand the ink style and artistic conception, we introduced an ink style prediction mechanism for our discriminator, which enables the discriminator to accurately predict the style with the help of a pre-trained style encoder. We also designed our generator to receive multi-scale feature information from the feature pyramid network for detail reconstruction of ink wash painting. Experimental results and user study show that ink wash paintings generated by our network have higher realism and richer artistic conception than existing image generation methods.

A novel Physical Layer Security (PLS) framework is conceived for enhancing the security of wireless communication systems by exploiting multi-domain polarization in Multiple-Input Multiple-Output (MIMO) systems. We design a sophisticated key generation scheme based on multi-domain polarization and the corresponding receivers. An in-depth analysis of the system’s secrecy rate is provided, demonstrating the confidentiality of our approach in the presence of eavesdroppers having strong computational capabilities. More explicitly, our simulation results and theoretical analysis corroborate the advantages of the proposed scheme in terms of its bit error rate (BER), block error rate (BLER), and maximum achievable secrecy rate. Our findings indicate that the innovative PLS framework effectively enhances the security and reliability of wireless communication systems. For example, in a$4\times 4$MIMO setup, the proposed PLS strategy exhibits an improvement of 2dB compared to conventional MIMO, systems at a BLER of$2\cdot 10^{-5}$while the eavesdropper’s BLER reaches 1.

Despite being widely deployed in safety-critical applications such as autonomous driving and health care, deep neural networks (DNNs) still suffer from non-negligible reliability issues. Numerous works had reported that DNNs were vulnerable to either natural environmental noises or man-made adversarial noises. How to repair DNNs in deployment with noisy samples is a crucial topic for the robustness of neural networks. While many network repairing methods based on data argumentation and weight adjustment have been proposed, they require retraining and redeploying the whole model, which causes high overhead and is infeasible for varying faulty cases on different deployment environments. In this paper, we propose a novel network repairing framework called PatchNAS from the architecture perspective, where we freeze the pretrained DNNs and introduce a small patch network to deal with failure samples at runtime. PatchNAS introduces a novel network instrumentation method to determine the faulty stage of the network structure given the collected failure samples. A small patch network structure is searched unsupervisedly using neural architecture search (NAS) technique with data samples from deployment environment. The patch network repairs the DNNs by correcting the output feature maps of the faulty stage, which helps to maintain network performance on normal samples and enhance robustness in noisy environments. Extensive experiments based on several DNNs across 15 types of natural noises show that the proposed PatchNAS outperforms the state-of-the-arts with significant performance improvement as well as much lower deployment overhead.

Abstract Colouring cartoon sketches can help children develop their intellect and inspire their artistic creativity. Unlike photo colourization or anime line art colourization, cartoon sketch colourization is challenging due to the scarcity of texture information and the irregularity of the line structure, which is mainly reflected in the phenomenon of colour‐bleeding artifacts in generated images. We propose a colourization approach for cartoon sketches, which takes both sketches and colour hints as inputs to produce impressive images. To solve the problem of colour‐bleeding artifacts, we propose a multi‐discriminator colourization framework that introduces a texture discriminator in the conditional generative adversarial network (cGAN). Then we combined this framework with a pre‐trained auxiliary auto‐encoder, where an auxiliary feature loss is designed to further improve colour quality, and a condition input is introduced to increase the generalization ability over hand‐drawn sketches. We present both quantitative and qualitative evaluations, which prove the effectiveness of our proposed method. We test our method on sketches of varying complexity and structure, then build an interactive programme based on our model for user study. Experimental results demonstrate that the method generates natural and consistent colour images in real time from sketches drawn by non‐professionals.

Owing to the rapid development of unmanned aerial vehicles (UAVs), UAV-enabled data collection has emerged as a promising technology. However, for the scenarios where the UAVs are dispatched to gather surrounding information actively and dynamically, the existing data collection methods can hardly fulfill the corresponding demands. To this end, we aim to study the paradigm data acquisition, where the UAV dynamically gathers information by on-board sensors. In this paper, we consider the scenario where a UAV acquires data in real time, which needs to be timely processed with the assistance of a mobile edge computing server. To address the real-time data acquisition characteristics, we construct a novel data acquisition rate model, which is with respect to the UAV speed. Next, we formulate a UAV energy consumption minimization problem that jointly considers UAV trajectory, transmission power, and CPU frequency. To tackle the highly complex problem, we propose an efficient iterative algorithm, and rigorously derive a closed-form solution for the UAV transmission power and computation resources allocation. With the obtained numerical results, we further validate the superiority of the proposed system design and the effectiveness of our algorithm against benchmark schemes.

Deep Convolutional Neural Networks (CNNs) are successfully applied in many complex tasks, but their storage and huge computational costs hinder their deployment on edge devices. CNN model compression techniques have been widely studied in the past five years, most of which are conducted in the spatial domain. Inspired by the sparsity and low-rank properties of weight matrices in the frequency domain, we propose a novel frequency pruning framework for model compression and acceleration while maintaining high-performance. We firstly apply Discrete Cosine Transform (DCT) on convolutional kernels and train them in the frequency domain to get sparse representations. Then we propose an iterative model compression method to decompose the frequency matrices with a sampled-based low-rank approximation algorithm, and then fine-tune and recompose the low-rank matrices gradually until a predefined compression ratio is reached. We further demonstrate that model inference can be conducted with the decomposed frequency matrices, where model parameters and inference cost can be significantly reduced. Extensive experiments using well-known CNN models based on three open datasets show that the proposed method outperforms the state-of-the-arts in reduction of both the number of parameters and floating-point operations (FLOPs) without sacrificing too much model accuracy.

Despite their great success in various fields, modern convolutional neural networks (CNNs) require huge amount of computation in inference due to their deeper network structure, which prevents them from being used in resource-limited devices such as mobile phones and embedded sensors. Recently, filter pruning had been introduced as a promising model compression method to reduce computation cost and storage overhead. However, existing filter pruning approaches are mainly model-based, which rely on empirical model to evaluate the importance of filters and set parameters manually to guide model compression. In this paper, we observe that CNNs commonly consist of large amount of inactive filters, and introduce Filter Inactive RatE (FIRE), a novel metric to evaluate the importance of filters in a neural network. Based on FIRE, we develop a learning based filter pruning strategy called FIREPruning for fast model compression. It adopts a regression model to predict the FIRE value and uses a three stage pipeline (FIRE prediction, pruning, and fine-tuning) to compress the neural network efficiently. Extensive experiments based on widely-used CNN models and well-known datasets show that FIREPruning reduces overall computation cost up to 86.9% without sacrificing too much accuracy, which significantly outperforms the state-of-the-art model compression methods.