Search

Formulated as a conditional generation problem, face animation aims at synthesizing continuous face images from a single source image driven by a set of conditional face motion. Previous works mainly model the face motion as conditions with 1D or 2D representation (e.g., action units, emotion codes, landmark), which often leads to low-quality results in some complicated scenarios such as continuous generation and largepose transformation. To tackle this problem, the conditions are supposed to meet two requirements, i.e., motion information preserving and geometric continuity. To this end, we propose a novel representation based on a 3D geometric flow, termed facial flow, to represent the natural motion of the human face at any pose. Compared with other previous conditions, the proposed facial flow well controls the continuous changes to the face. After that, in order to utilize the facial flow for face editing, we build a synthesis framework generating continuous images with conditional facial flows. To fully take advantage of the motion information of facial flows, a hierarchical conditional framework is designed to combine the extracted multi-scale appearance features from images and motion features from flows in a hierarchical manner. The framework then decodes multiple fused features back to images progressively. Experimental results demonstrate the effectiveness of our method compared to other state-of-the-art methods.

This paper Initial on the solar parabolic trough collectors…Semi-insulated voltage transformers are widely used in power distribution network. The safety of their own equipment directly affects the safe and stable operation of power system. In this paper, the fault diagnosis and analysis process of a semi-insulated voltage transformer is described in detail. In this paper, the fault course and the condition of the fault equipment are introduced at first, and then the cause of the fault is clarified through the electrical test and disassembly analysis combined with the action protection condition. Finally, the preventive measures for the daily operation and maintenance of the equipment are put forward.

Deep Learning has implemented a wide range of applications and has become increasingly popular in recent years. The goal of multimodal deep learning is to create models that can process and link information using various modalities. Despite the extensive development made for unimodal learning, it still cannot cover all the aspects of human learning. Multimodal learning helps to understand and analyze better when various senses are engaged in the processing of information. This paper focuses on multiple types of modalities, i.e., image, video, text, audio, body gestures, facial expressions, and physiological signals. Detailed analysis of past and current baseline approaches and an in-depth study of recent advancements in multimodal deep learning applications has been provided. A fine-grained taxonomy of various multimodal deep learning applications is proposed, elaborating on different applications in more depth. Architectures and datasets used in these applications are also discussed, along with their evaluation metrics. Last, main issues are highlighted separately for each domain along with their possible future research directions.

Recent approaches for fast semantic video segmentation have reduced redundancy by warping feature maps across adjacent frames, greatly speeding up the inference phase. However, the accuracy drops seriously owing to the errors incurred by warping. In this paper, we propose a novel framework and design a simple and effective correction stage after warping. Specifically, we build a non-key-frame CNN, fusing warped context features with current spatial details. Based on the feature fusion, our Context Feature Rectification~(CFR) module learns the model's difference from a per-frame model to correct the warped features. Furthermore, our Residual-Guided Attention~(RGA) module utilizes the residual maps in the compressed domain to help CRF focus on error-prone regions. Results on Cityscapes show that the accuracy significantly increases from $67.3\%$ to $71.6\%$, and the speed edges down from $65.5$ FPS to $61.8$ FPS at a resolution of $1024\times 2048$. For non-rigid categories, e.g., ``human'' and ``object'', the improvements are even higher than 18 percentage points.

The study of generative adversarial networks (GAN) provides a new approach and framework for computer vision and makes an outstanding choice for cross-domain image translation problems. In this work, the basic framework structure of GAN as well as the evolution of GAN are summarized. It is known that the unsupervised domain adaptation algorithms attempt to map representations between the two domains, or learn to extract features that are domain–invariant. An approach that learns in an unsupervised manner can be a transformation in the pixel space from one domain to the other. Finally, some existing problems of GAN and its potential future are also summarized and discussed.

Dry-type iron core reactors frequently have various insulation faults due to their own process defects and operating environment impacts, which pose a certain threat to the safe and stable operation of the power grid. In this paper, the technicians conducted overall tests, on-site analysis, and comprehensive judgment on a 220kV dry-type iron-core shunt reactor, and proposed solutions on this basis, which have certain guiding significance for solving the insulation failure of the iron-core shunt reactor.

Prokaryotic restriction enzymes, recombinases and Cas proteins are powerful DNA engineering and genome editing tools. However, in many primary cell types, the efficiency of genome editing remains low, impeding the development of gene- and cell-based therapeutic applications. A safe strategy for robust and efficient enrichment of precisely genetically engineered cells is urgently required. Here, we screen for mutations in the receptor for Diphtheria Toxin (DT) which protect human cells from DT. Selection for cells with an edited DT receptor variant enriches for simultaneously introduced, precisely targeted gene modifications at a second independent locus, such as nucleotide substitutions and DNA insertions. Our method enables the rapid generation of a homogenous cell population with bi-allelic integration of a DNA cassette at the selection locus, without clonal isolation. Toxin-based selection works in both cancer-transformed and non-transformed cells, including human induced pluripotent stem cells and human primary T-lymphocytes, as well as it is applicable also in vivo, in mice with humanized liver. This work represents a flexible, precise, and efficient selection strategy to engineer cells using CRISPR-Cas and base editing systems., Genome engineering in cell lines or human stem cells often has poor efficiency, limiting the development of research and therapeutic applications. Here, the authors use a toxin-based selection system for precise bi-allelic engineering in cells and in vivo.