arXiv subjects

Wei Zhou

Publications and source records attributed to Wei Zhou.

At least 361 records · Page 20Linked to original sources

arXiv2022

No-Reference Light Field Image Quality Assessment Based on Spatial-Angular Measurement

Light field image quality assessment (LFI-QA) is a significant and challenging research problem. It helps to better guide light field acquisition, processing and applications. However, only a few objective models have been proposed and none of them completely consider intrinsic factors affecting the LFI quality. In this paper, we propose a No-Reference Light Field image Quality Assessment (NR-LFQA) scheme, where the main idea is to quantify the LFI quality degradation through evaluating the spatial quality and angular consistency. We first measure the spatial quality deterioration by capturing the naturalness distribution of the light field cyclopean image array, which is formed when human observes the LFI. Then, as a transformed representation of LFI, the Epipolar Plane Image (EPI) contains the slopes of lines and involves the angular information. Therefore, EPI is utilized to extract the global and local features from LFI to measure angular consistency degradation. Specifically, the distribution of gradient direction map of EPI is proposed to measure the global angular consistency distortion in the LFI. We further propose the weighted local binary pattern to capture the characteristics of local angular consistency degradation. Extensive experimental results on four publicly available LFI quality datasets demonstrate that the proposed method outperforms state-of-the-art 2D, 3D, multi-view, and LFI quality assessment algorithms.

eess.IV↗

arXiv2022

Tensor Oriented No-Reference Light Field Image Quality Assessment

Light field image (LFI) quality assessment is becoming more and more important, which helps to better guide the acquisition, processing and application of immersive media. However, due to the inherent high dimensional characteristics of LFI, the LFI quality assessment turns into a multi-dimensional problem that requires consideration of the quality degradation in both spatial and angular dimensions. Therefore, we propose a novel Tensor oriented No-reference Light Field image Quality evaluator (Tensor-NLFQ) based on tensor theory. Specifically, since the LFI is regarded as a low-rank 4D tensor, the principal components of four oriented sub-aperture view stacks are obtained via Tucker decomposition. Then, the Principal Component Spatial Characteristic (PCSC) is designed to measure the spatial-dimensional quality of LFI considering its global naturalness and local frequency properties. Finally, the Tensor Angular Variation Index (TAVI) is proposed to measure angular consistency quality by analyzing the structural similarity distribution between the first principal component and each view in the view stack. Extensive experimental results on four publicly available LFI quality databases demonstrate that the proposed Tensor-NLFQ model outperforms state-of-the-art 2D, 3D, multi-view, and LFI quality assessment algorithms.

eess.IV↗

arXiv2022

On Language Model Integration for RNN Transducer based Speech Recognition

The mismatch between an external language model (LM) and the implicitly learned internal LM (ILM) of RNN-Transducer (RNN-T) can limit the performance of LM integration such as simple shallow fusion. A Bayesian interpretation suggests to remove this sequence prior as ILM correction. In this work, we study various ILM correction-based LM integration methods formulated in a common RNN-T framework. We provide a decoding interpretation on two major reasons for performance improvement with ILM correction, which is further experimentally verified with detailed analysis. We also propose an exact-ILM training framework by extending the proof given in the hybrid autoregressive transducer, which enables a theoretical justification for other ILM approaches. Systematic comparison is conducted for both in-domain and cross-domain evaluation on the Librispeech and TED-LIUM Release 2 corpora, respectively. Our proposed exact-ILM training can further improve the best ILM method.

cs.CL↗

arXiv2022

Topmetal-M: a novel pixel sensor for compact tracking applications

The Topmetal-M is a large area pixel sensor (18 mm * 23 mm) prototype fabricated in a new 130 nm high-resistivity CMOS process in 2019. It contains 400 rows * 512 columns square pixels with the pitch of 40 μm. In Topmetal-M, a novel charge collection method combing the Monolithic Active Pixel Sensor (MAPS) and the Topmetal sensor has been proposed for the first time. Both the ionized charge deposited by the particle in the sensor and along the track over the sensor can be collected. The in-pixel circuit mainly consists of a low-noise charge sensitive amplifier to establish the signal for the energy reconstruction, and a discriminator with a Time-to-Amplitude Converter (TAC) for the Time of Arrival (TOA) measurement. With this mechanism, the trajectory, particle hit position, energy and arrival time of the particle can be measured. The analog signal from each pixel is accessible through time-shared multiplexing over the entire pixel array. This paper will discuss the design and preliminary test results of the Topmetal-M sensor.

physics.ins-det↗

arXiv2021

Maxwell: a hardware and software highly integrated compute-storage system

The compute-storage framework is responsible for data storage and processing, and acts as the digital chassis of all upper-level businesses. The performance of the framework affects the business's processing throughput, latency, jitter, and etc., and also determines the theoretical performance upper bound that the business can achieve. In financial applications, the compute-storage framework must have high reliability and high throughput, but with low latency as well as low jitter characteristics. For some scenarios such as hot-spot account update, the performance of the compute-storage framework even surfaces to become a server performance bottleneck of the whole business system. In this paper, we study the hot-spot account issue faced by Alipay and present our exciting solution to this problem by developing a new compute-storage system, called Maxwell. Maxwell is a distributed compute-storage system with integrated hardware and software optimizations. Maxwell does not rely on any specific hardware (e.g. GPUs or FPGAs). Instead, it takes deep advantage of computer components' characteristics, such as disk, network, operating system and CPU, and aims to emit the ultimate performance of both hardware and software. In comparison with the existing hot-spot account updating solutions deployed online, Maxwell achieves three orders of magnitude performance improvement for end-to-end evaluation. Meanwhile, Maxwell also demonstrates remarkable performance gains in other related businesses of Ant Group.

cs.DC↗

arXiv2021

Raw Bayer Pattern Image Synthesis for Computer Vision-oriented Image Signal Processing Pipeline Design

In this paper, we propose a method to add constraints that are un-formulatable in generative adversarial networks (GAN)-based arbitrary size RAW Bayer image generation. It is shown theoretically that by using the transformed data in GAN training, it is able to improve the learning of the original data distribution, owing to the invariant of Jensen-Shannon (JS) divergence between two distributions under invertible and differentiable transformation. Benefiting from the proposed method, RAW Bayer pattern images can be generated by configuring the transformation as demosaicing. It is shown that by adding another transformation, the proposed method is able to synthesize high-quality RAW Bayer images with arbitrary size. Experimental results show that images generated by the proposed method outperform the existing methods in the Fréchet inception distance (FID) score, peak signal to noise ratio (PSNR), and mean structural similarity (MSSIM), and the training process is more stable. To the best knowledge of the authors, there is no open-source, large-scale image dataset in the RAW Bayer domain, which is crucial for research works aiming to explore the image signal processing (ISP) pipeline design for computer vision tasks. Converting the existing commonly used color image datasets to their corresponding RAW Bayer versions, the proposed method can be a promising solution to the RAW image dataset problem. We also show in the experiments that, by training object detection frameworks using the synthesized RAW Bayer images, they can be used in an end-to-end manner (from RAW images to vision tasks) with negligible performance degradation.

eess.IV↗

arXiv2021

ALOHA: Artificial Learning of Human Attributes for Dialogue Agents

For conversational AI and virtual assistants to communicate with humans in a realistic way, they must exhibit human characteristics such as expression of emotion and personality. Current attempts toward constructing human-like dialogue agents have presented significant difficulties. We propose Human Level Attributes (HLAs) based on tropes as the basis of a method for learning dialogue agents that can imitate the personalities of fictional characters. Tropes are characteristics of fictional personalities that are observed recurrently and determined by viewers' impressions. By combining detailed HLA data with dialogue data for specific characters, we present a dataset, HLA-Chat, that models character profiles and gives dialogue agents the ability to learn characters' language styles through their HLAs. We then introduce a three-component system, ALOHA (which stands for Artificial Learning of Human Attributes), that combines character space mapping, character community detection, and language style retrieval to build a character (or personality) specific language model. Our preliminary experiments demonstrate that two variations of ALOHA, combined with our proposed dataset, can outperform baseline models at identifying the correct dialogue responses of chosen target characters, and are stable regardless of the character's identity, the genre of the show, and the context of the dialogue.

cs.CL↗

arXiv2021

Prediction of Listener Perception of Argumentative Speech in a Crowdsourced Dataset Using (Psycho-)Linguistic and Fluency Features

One of the key communicative competencies is the ability to maintain fluency in monologic speech and the ability to produce sophisticated language to argue a position convincingly. In this paper we aim to predict TED talk-style affective ratings in a crowdsourced dataset of argumentative speech consisting of 7 hours of speech from 110 individuals. The speech samples were elicited through task prompts relating to three debating topics. The samples received a total of 2211 ratings from 737 human raters pertaining to 14 affective categories. We present an effective approach to the classification task of predicting these categories through fine-tuning a model pre-trained on a large dataset of TED talks public speeches. We use a combination of fluency features derived from a state-of-the-art automatic speech recognition system and a large set of human-interpretable linguistic features obtained from an automatic text analysis system. Classification accuracy was greater than 60% for all 14 rating categories, with a peak performance of 72% for the rating category 'informative'. In a secondary experiment, we determined the relative importance of features from different groups using SP-LIME.

cs.CL↗

arXiv2021

Efficient Learning of Quadratic Variance Function Directed Acyclic Graphs via Topological Layers

Directed acyclic graph (DAG) models are widely used to represent causal relationships among random variables in many application domains. This paper studies a special class of non-Gaussian DAG models, where the conditional variance of each node given its parents is a quadratic function of its conditional mean. Such a class of non-Gaussian DAG models are fairly flexible and admit many popular distributions as special cases, including Poisson, Binomial, Geometric, Exponential, and Gamma. To facilitate learning, we introduce a novel concept of topological layers, and develop an efficient DAG learning algorithm. It first reconstructs the topological layers in a hierarchical fashion and then recoveries the directed edges between nodes in different layers, which requires much less computational cost than most existing algorithms in literature. Its advantage is also demonstrated in a number of simulated examples, as well as its applications to two real-life datasets, including an NBA player statistics data and a cosmetic sales data collected by Alibaba.

stat.ML↗

arXiv2021

Tunable Magnetic Skyrmions in Ferrimagnetic Mn$_4$N

Thin films of ferrimagnetic Mn$_4$N are candidate materials to host magnetic skyrmions that have demonstrated thermal stability up to 450$^\circ$C. However, there are no experimental reports observing skyrmions in this system. Here, we discuss the results of sputter grown 15-17~nm Mn$_4$N thin films on MgO substrate capped with Pt$_{1-x}$Cu$_{x}$ layers. Vibrating sample magnetometry measurement of out-of-plane hysteresis loops confirmed that magnetic properties are insensitive to the cap layer composition. Imaging based on magnetic force microscopy measurements observed 300 to 50~nm sized skyrmions, as the Cu concentration was increased from $x$ = 0 to 0.9. We performed density functional theory calculations and found that the interfacial Dzyaloshinskii-Moriya interactions (iDMI) follow a trend: Mn$_4$N/MgO(001) $<$ Cu/Mn$_4$N(001) $<$ Pt/Mn$_4$N(001). We infer from these calculations that $x$ in Pt$_{1-x}$Cu$_{x}$ capping layer can serve as a robust tuning knob to tailor the iDMI and control the skyrmion size. This work provides guidance to achieve smaller Néel-type skyrmions in Mn$_4$N thin films, which is an important step forward for building thermally stable skyrmionic devices.

cond-mat.mes-hall↗

arXiv2021

Gradient-based Feature Extraction From Raw Bayer Pattern Images

In this paper, the impact of demosaicing on gradient extraction is studied and a gradient-based feature extraction pipeline based on raw Bayer pattern images is proposed. It is shown both theoretically and experimentally that the Bayer pattern images are applicable to the central difference gradient-based feature extraction algorithms without performance degradation, or even with superior performance in some cases. The color difference constancy assumption, which is widely used in various demosaicing algorithms, is applied in the proposed Bayer pattern image-based gradient extraction pipeline. Experimental results show that the gradients extracted from Bayer pattern images are robust enough to be used in histogram of oriented gradients (HOG)-based pedestrian detection algorithms and shift-invariant feature transform (SIFT)-based matching algorithms.

eess.IV↗

arXiv2021

Large-Gap Quantum Spin Hall State and Temperature-Induced Lifshitz Transition in Bi4Br4

Searching for new quantum spin Hall insulators with large fully opened energy gap to overcome the thermal disturbance at room temperature has attracted tremendous attention due to the one-dimensional (1D) spin-momentum locked topological edge states serving as dissipationless channels for the practical applications in low consumption electronics and high performance spintronics. Here, we report the investigation of topological nature of monolayer Bi4Br4 by the techniques of scanning tunneling microscopy and angle-resolved photoemission spectroscopy (ARPES). The topological non-triviality of 1D edge state integrals within the large bulk energy gap (~ 0.2 eV) is revealed by the first-principle calculations. The ARPES measurements at different temperature show a temperature-induced Lifshitz transition, corresponding to the resistivity anomaly caused by the shift of chemical potential. The connection between the emergency of superconductivity and the Lifshitz transition is discussed.

cond-mat.mtrl-sci↗

arXiv2021

AliMe MKG: A Multi-modal Knowledge Graph for Live-streaming E-commerce

Live streaming is becoming an increasingly popular trend of sales in E-commerce. The core of live-streaming sales is to encourage customers to purchase in an online broadcasting room. To enable customers to better understand a product without jumping out, we propose AliMe MKG, a multi-modal knowledge graph that aims at providing a cognitive profile for products, through which customers are able to seek information about and understand a product. Based on the MKG, we build an online live assistant that highlights product search, product exhibition and question answering, allowing customers to skim over item list, view item details, and ask item-related questions. Our system has been launched online in the Taobao app, and currently serves hundreds of thousands of customers per day.

cs.IR↗

arXiv2021

DLB: Deep Learning Based Load Balancing

In this paper, we introduce DLB, a Deep Learning based load Balancing mechanism, to effectively address the data skew problem. The key idea of DLB is to replace hash functions in the load balancing mechanisms with deep learning models, which are trained to be able to map different distributions of workloads and data to the servers in a uniform manner. We implemented DLB and deployed it on a practical Cloud environment using CloudSim. Experimental results using both synthetic and real-world data sets show that compared with traditional hash function-based load balancing methods, DLB is able to achieve more balanced mappings, especially when the workload is highly skewed.

cs.DC↗

arXiv2021

Acoustic Data-Driven Subword Modeling for End-to-End Speech Recognition

Subword units are commonly used for end-to-end automatic speech recognition (ASR), while a fully acoustic-oriented subword modeling approach is somewhat missing. We propose an acoustic data-driven subword modeling (ADSM) approach that adapts the advantages of several text-based and acoustic-based subword methods into one pipeline. With a fully acoustic-oriented label design and learning process, ADSM produces acoustic-structured subword units and acoustic-matched target sequence for further ASR training. The obtained ADSM labels are evaluated with different end-to-end ASR approaches including CTC, RNN-Transducer and attention models. Experiments on the LibriSpeech corpus show that ADSM clearly outperforms both byte pair encoding (BPE) and pronunciation-assisted subword modeling (PASM) in all cases. Detailed analysis shows that ADSM achieves acoustically more logical word segmentation and more balanced sequence length, and thus, is suitable for both time-synchronous and label-synchronous models. We also briefly describe how to apply acoustic-based subword regularization and unseen text segmentation using ADSM.

cs.CL↗

arXiv2021

SPMoE: Generate Multiple Pattern-Aware Outputs with Sparse Pattern Mixture of Experts

Many generation tasks follow a one-to-many mapping relationship: each input could be associated with multiple outputs. Existing methods like Conditional Variational AutoEncoder(CVAE) employ a latent variable to model this one-to-many relationship. However, this high-dimensional and dense latent variable lacks explainability and usually leads to poor and uncontrollable generations. In this paper, we innovatively introduce the linguistic concept of pattern to decompose the one-to-many mapping into multiple one-to-one mappings and further propose a model named Sparse Pattern Mixture of Experts(SPMoE). Each one-to-one mapping is associated with a conditional generation pattern and is modeled with an expert in SPMoE. To ensure each language pattern can be exclusively handled with an expert model for better explainability and diversity, a sparse mechanism is employed to coordinate all the expert models in SPMoE. We assess the performance of our SPMoE on the paraphrase generation task and the experiment results prove that SPMoE can achieve a good balance in terms of quality, pattern-level diversity, and corpus-level diversity.

cs.CL↗

arXiv2021

GGP: A Graph-based Grouping Planner for Explicit Control of Long Text Generation

Existing data-driven methods can well handle short text generation. However, when applied to the long-text generation scenarios such as story generation or advertising text generation in the commercial scenario, these methods may generate illogical and uncontrollable texts. To address these aforementioned issues, we propose a graph-based grouping planner(GGP) following the idea of first-plan-then-generate. Specifically, given a collection of key phrases, GGP firstly encodes these phrases into an instance-level sequential representation and a corpus-level graph-based representation separately. With these two synergic representations, we then regroup these phrases into a fine-grained plan, based on which we generate the final long text. We conduct our experiments on three long text generation datasets and the experimental results reveal that GGP significantly outperforms baselines, which proves that GGP can control the long text generation by knowing how to say and in what order.

cs.CL↗

arXiv2021

Transformer-Encoder-GRU (T-E-GRU) for Chinese Sentiment Analysis on Chinese Comment Text

Chinese sentiment analysis (CSA) has always been one of the challenges in natural language processing due to its complexity and uncertainty. Transformer has succeeded in capturing semantic features, but it uses position encoding to capture sequence features, which has great shortcomings compared with the recurrent model. In this paper, we propose T-E-GRU for Chinese sentiment analysis, which combine transformer encoder and GRU. We conducted experiments on three Chinese comment datasets. In view of the confusion of punctuation marks in Chinese comment texts, we selectively retain some punctuation marks with sentence segmentation ability. The experimental results show that T-E-GRU outperforms classic recurrent model and recurrent model with attention.

cs.CL↗