Considerable progress has recently been made in leveraging CLIP (Contrastive Language-Image Pre-Training) models for text-guided image manipulation. However, all existing works rely on additional generative models to ensure the quality of results, because CLIP alone cannot provide enough guidance information for fine-scale pixel-level changes. In this paper, we introduce CLIPVG, a text-guided image manipulation framework using differentiable vector graphics, which is also the first CLIP-based general image manipulation framework that does not require any additional generative models. We demonstrate that CLIPVG can not only achieve state-of-art performance in both semantic correctness and synthesis quality, but also is flexible enough to support various applications far beyond the capability of all existing methods.

For unsupervised image-to-image translation, we propose a discriminator architecture which focuses on the statistical features instead of individual patches. The network is stabilized by distribution matching of key statistical features at multiple scales. Unlike the existing methods which impose more and more constraints on the generator, our method facilitates the shape deformation and enhances the fine details with a greatly simplified framework. We show that the proposed method outperforms the existing state-of-the-art models in various challenging applications including selfie-to-anime, male-to-female and glasses removal.

3GPP has completed the first phase of the standardization of 5G cellular systems covering bands up to 52.6 GHz. The 3GPP Rel-15 New Radio standard provides a comprehensive framework for supporting massive MIMO in a wide variety of use cases and deployment scenarios. The NR-MIMO framework supports large scale antenna arrays having arbitrary configurations and supports digital, hybrid, and analog array architectures for both FDD and TDD deployments. With the rollout of NR soon to commence, many questions have been raised on the benefits and performance of NR-MIMO especially compared to the FD-MIMO framework in Rel-13/14 of LTE. This paper provides detailed system level simulation results showing the performance characteristics of NR-MIMO as a function of a various system and deployment parameters such as the array configuration, deployment scenario, and transmission scheme. The purpose is to answer important questions for network planners looking to roll-out NR-MIMO in the near future.

In this paper, we propose a link outage probability based resource allocation scheme for multi-radio multi-channel wireless mesh networks. The objective is to optimize the link outage probability under the effect of channel variation and external interference while preserving the end-to-end traffic demand requirements. The problem is formulated as a mixed-integer nonlinear programming problem, which is solved by decomposing it into a feasibility- checking problem and an outage probability search problem so that the original problem can be solved iteratively with reduced complexity. Numerical results are provided to show that the proposed algorithm is superior to the existing scheme in reducing the outage probability.

In this paper, we consider the robust resource provisioning problem targeting for end-to-end rate guarantee under moderate channel variations and external interference. We formulate the robustness optimization problem that explicitly takes into account practical radio switching, co-channel contention, and multi-path routing and traffic splitting constraints. By exploiting the special property of the problem, we propose a scheme to decompose the problem into a feasibility-checking problem and an interference margin search problem, which is guaranteed to converge to the optimal value with reduced complexity. Using traces collected from an indoor wireless network testbed, we evaluate the performance of the proposed algorithm, and show that the algorithm is superior to existing scheme in providing larger interference margin and reducing packet loss probability.