arXiv subjects

Tao Li

Publications and source records attributed to Tao Li.

At least 379 records · Page 21Linked to original sources

arXiv2022

Cross-speaker Emotion Transfer Based On Prosody Compensation for End-to-End Speech Synthesis

Cross-speaker emotion transfer speech synthesis aims to synthesize emotional speech for a target speaker by transferring the emotion from reference speech recorded by another (source) speaker. In this task, extracting speaker-independent emotion embedding from reference speech plays an important role. However, the emotional information conveyed by such emotion embedding tends to be weakened in the process to squeeze out the source speaker's timbre information. In response to this problem, a prosody compensation module (PCM) is proposed in this paper to compensate for the emotional information loss. Specifically, the PCM tries to obtain speaker-independent emotional information from the intermediate feature of a pre-trained ASR model. To this end, a prosody compensation encoder with global context (GC) blocks is introduced to obtain global emotional information from the ASR model's intermediate feature. Experiments demonstrate that the proposed PCM can effectively compensate the emotion embedding for the emotional information loss, and meanwhile maintain the timbre of the target speaker. Comparisons with state-of-the-art models show that our proposed method presents obvious superiority on the cross-speaker emotion transfer task.

cs.SD↗

arXiv2022

Heegaard genus, degree-one maps, and amalgamation of 3-manifolds

Let $M=W\cup_T V$ be an amalgamation of two compact 3-manifolds along a torus, where $W$ is the exterior of a knot in a homology sphere. Let $N$ be the manifold obtained by replacing $W$ with a solid torus such that the boundary of a Seifert surface in $W$ is a meridian of the solid torus. This means that there is a degree-one map $f\colon M\to N$, pinching $W$ into a solid torus while fixing $V$. We prove that $g(M)\ge g(N)$, where $g(M)$ denotes the Heegaard genus. An immediate corollary is that the tunnel number of a satellite knot is at least as large as the tunnel number of its pattern knot.

math.GT↗

arXiv2022

Progressive Multi-scale Consistent Network for Multi-class Fundus Lesion Segmentation

Effectively integrating multi-scale information is of considerable significance for the challenging multi-class segmentation of fundus lesions because different lesions vary significantly in scales and shapes. Several methods have been proposed to successfully handle the multi-scale object segmentation. However, two issues are not considered in previous studies. The first is the lack of interaction between adjacent feature levels, and this will lead to the deviation of high-level features from low-level features and the loss of detailed cues. The second is the conflict between the low-level and high-level features, this occurs because they learn different scales of features, thereby confusing the model and decreasing the accuracy of the final prediction. In this paper, we propose a progressive multi-scale consistent network (PMCNet) that integrates the proposed progressive feature fusion (PFF) block and dynamic attention block (DAB) to address the aforementioned issues. Specifically, PFF block progressively integrates multi-scale features from adjacent encoding layers, facilitating feature learning of each layer by aggregating fine-grained details and high-level semantics. As features at different scales should be consistent, DAB is designed to dynamically learn the attentive cues from the fused features at different scales, thus aiming to smooth the essential conflicts existing in multi-scale features. The two proposed PFF and DAB blocks can be integrated with the off-the-shelf backbone networks to address the two issues of multi-scale and feature inconsistency in the multi-class segmentation of fundus lesions, which will produce better feature representation in the feature space. Experimental results on three public datasets indicate that the proposed method is more effective than recent state-of-the-art methods.

eess.IV↗

arXiv2022

Secure and Efficient Tunneling of MACsec for Modern Industrial Use Cases

Trends like Industry 4.0 will pose new challenges for future industrial networks. Greater interconnectedness, higher data volumes as well as new requirements for speeds as well as security will make new approaches necessary. Performanceoptimized networking techniques will be demanded to implement new use cases, like network separation and isolation, in a secure fashion. A new and highly efficient protocol, that will be vital for that purpose, is MACsec. It is a Layer 2 encryption protocol that was previously extended specifically for industrial environments. Yet, it lacks the ability to bridge local networks. Therefore, in this work, we propose a secure and efficient Layer 3 tunneling scheme for MACsec. We design and implement two approaches, that are equally secure and considerably outperform comparable state-of-the-art techniques.

cs.CR↗

arXiv2022

MSTRIQ: No Reference Image Quality Assessment Based on Swin Transformer with Multi-Stage Fusion

Measuring the perceptual quality of images automatically is an essential task in the area of computer vision, as degradations on image quality can exist in many processes from image acquisition, transmission to enhancing. Many Image Quality Assessment(IQA) algorithms have been designed to tackle this problem. However, it still remains un settled due to the various types of image distortions and the lack of large-scale human-rated datasets. In this paper, we propose a novel algorithm based on the Swin Transformer [31] with fused features from multiple stages, which aggregates information from both local and global features to better predict the quality. To address the issues of small-scale datasets, relative rankings of images have been taken into account together with regression loss to simultaneously optimize the model. Furthermore, effective data augmentation strategies are also used to improve the performance. In comparisons with previous works, experiments are carried out on two standard IQA datasets and a challenge dataset. The results demonstrate the effectiveness of our work. The proposed method outperforms other methods on standard datasets and ranks 2nd in the no-reference track of NTIRE 2022 Perceptual Image Quality Assessment Challenge [53]. It verifies that our method is promising in solving diverse IQA problems and thus can be used to real-word applications.

cs.CV↗

arXiv2022

Cross-speaker emotion disentangling and transfer for end-to-end speech synthesis

The cross-speaker emotion transfer task in text-to-speech (TTS) synthesis particularly aims to synthesize speech for a target speaker with the emotion transferred from reference speech recorded by another (source) speaker. During the emotion transfer process, the identity information of the source speaker could also affect the synthesized results, resulting in the issue of speaker leakage. This paper proposes a new method with the aim to synthesize controllable emotional expressive speech and meanwhile maintain the target speaker's identity in the cross-speaker emotion TTS task. The proposed method is a Tacotron2-based framework with emotion embedding as the conditioning variable to provide emotion information. Two emotion disentangling modules are contained in our method to 1) get speaker-irrelevant and emotion-discriminative embedding, and 2) explicitly constrain the emotion and speaker identity of synthetic speech to be that as expected. Moreover, we present an intuitive method to control the emotion strength in the synthetic speech for the target speaker. Specifically, the learned emotion embedding is adjusted with a flexible scalar value, which allows controlling the emotion strength conveyed by the embedding. Extensive experiments have been conducted on a Mandarin disjoint corpus, and the results demonstrate that the proposed method is able to synthesize reasonable emotional speech for the target speaker. Compared to the state-of-the-art reference embedding learned methods, our method gets the best performance on the cross-speaker emotion transfer task, indicating that our method achieves the new state-of-the-art performance on learning the speaker-irrelevant emotion embedding. Furthermore, the strength ranking test and pitch trajectories plots demonstrate that the proposed method can effectively control the emotion strength, leading to prosody-diverse synthetic speech.

cs.SD↗

arXiv2022

Subspace Adversarial Training

Single-step adversarial training (AT) has received wide attention as it proved to be both efficient and robust. However, a serious problem of catastrophic overfitting exists, i.e., the robust accuracy against projected gradient descent (PGD) attack suddenly drops to 0% during the training. In this paper, we approach this problem from a novel perspective of optimization and firstly reveal the close link between the fast-growing gradient of each sample and overfitting, which can also be applied to understand robust overfitting in multi-step AT. To control the growth of the gradient, we propose a new AT method, Subspace Adversarial Training (Sub-AT), which constrains AT in a carefully extracted subspace. It successfully resolves both kinds of overfitting and significantly boosts the robustness. In subspace, we also allow single-step AT with larger steps and larger radius, further improving the robustness performance. As a result, we achieve state-of-the-art single-step AT performance. Without any regularization term, our single-step AT can reach over 51% robust accuracy against strong PGD-50 attack of radius 8/255 on CIFAR-10, reaching a competitive performance against standard multi-step PGD-10 AT with huge computational advantages. The code is released at https://github.com/nblt/Sub-AT.

cs.LG↗

arXiv2022

Competing emergent Potts orders and possible nematic spin liquids in the Kagome $J_{1}-J_{3}$ Heisenberg model

The extreme frustration in the Kagome antiferromagnet makes it possible to realize a large number of closely competing magnetic phases either at the classical or the quantum level. Motivated by recent neutron scattering study on the Kagome antiferromagnet vesignieite BaCu$_{3}$V$_{2}$O$_{8}$(OH)$_{2}$, we have made systematic investigation of the phase diagram of the classical and quantum $J_{1}-J_{3}$ antiferromagnetic Heisenberg model defined on the Kagome lattice(KAFHM). While it is shown previously that a large antiferromagnetic exchange between the third-neighboring spins can drive an emergent $q=4$ Potts order through the order-by-disorder mechanism, it is elusive what will happen in the so called "grey region" where the Luttinger-Tisza criteria fails to predict the classical ordering pattern. Through extensive Monte Carlo simulation, we find that the "grey region" is characterized by competing emergent $q=3$ and $q=4$ Potts order, whose emergence are beyond the description of the conventional order-by-disorder mechanism. Our Schwinger Boson mean field calculation in the "grey region" confirms the existence of the Potts-3 order in the ground state of quantum $J_{1}-J_{3}$ KAFHM. The predicted Potts-3 state is found to take two different forms distinguished by their projective symmetry group character(PSG or quantum order), although they have exactly the same symmetry. A comparison with the exact diagonalization result on a 36-site cluster shows that the "grey region" may indeed host a nematic spin liquid ground state featuring an anisotropic ring structure around the $\mathbf{q}=0$ point in its spinon dispersion relation. We find that such "grey region" of extremely frustrated magnets should better be taken as the playground to study the rich competition between exotic emergent phases, rather than a burden on their theoretical analysis.

cond-mat.str-el↗

arXiv2022

Tightly Coupled Optimization-based GPS-Visual-Inertial Odometry with Online Calibration and Initialization

In this paper, we present a tightly coupled optimization-based GPS-Visual-Inertial odometry system to solve the trajectory drift of the visual-inertial odometry especially over long-term runs. Visual reprojection residuals, IMU residuals, and GPS measurement residuals are jointly minimized within a local bundle adjustment, in which we apply GPS measurements and IMU preintegration used for the IMU residuals to formulate a novel GPS residual. To improve the efficiency and robustness of the system, we propose a fast reference frames initialization method and an online calibration method for GPS-IMU extrinsic and time offset. In addition, we further test the performance and convergence of our online calibration method. Experimental results on EuRoC datasets show that our method consistently outperforms other tightly coupled and loosely coupled approaches. Meanwhile, this system has been validated on KAIST datasets, which proves that our system can work well in the case of visual or GPS failure.

cs.RO↗

arXiv2022

Measurement of high-pressure xenon gas absorption in acrylic

Acrylic is a popular structural material in experiments requiring low background because of its radio-purity, machinability, and mechanical strength. However, its porosity may cause significant gas absorption and influence the detector stability in the long term. The interaction between acrylic and other detector materials becomes one of the key concerns in the detector design. In this paper, we carry out an experiment to measure quantitatively the absorption process of high-pressure xenon gas into acrylic. A specific setup is designed for the measurement, and systematic measurements are done to obtain a result of the absorption amount: 0.98 g xenon into 332 g of acrylic.

physics.ins-det↗

arXiv2022

One-shot Voice Conversion For Style Transfer Based On Speaker Adaptation

One-shot style transfer is a challenging task, since training on one utterance makes model extremely easy to over-fit to training data and causes low speaker similarity and lack of expressiveness. In this paper, we build on the recognition-synthesis framework and propose a one-shot voice conversion approach for style transfer based on speaker adaptation. First, a speaker normalization module is adopted to remove speaker-related information in bottleneck features extracted by ASR. Second, we adopt weight regularization in the adaptation process to prevent over-fitting caused by using only one utterance from target speaker as training data. Finally, to comprehensively decouple the speech factors, i.e., content, speaker, style, and transfer source style to the target, a prosody module is used to extract prosody representation. Experiments show that our approach is superior to the state-of-the-art one-shot VC systems in terms of style and speaker similarity; additionally, our approach also maintains good speech quality.

eess.AS↗

arXiv2022

Sender-controlled measurement-device-independent multiparty quantum communication

Multiparty quantum communication is an important branch of quantum networks. It enables private information transmission with information-theoretic security among legitimate parties. We propose a sender-controlled measurement-device-independent multiparty quantum communication protocol. The sender Alice divides a private message into several parts and delivers them to different receivers for secret sharing with imperfect measurement devices and untrusted ancillary nodes. Furthermore, Alice acts as an active controller and checks the security of quantum channels and the reliability of each receiver before she encodes her private message for secret sharing, which makes the protocol convenient for multiparity quantum communication.

quant-ph↗

arXiv2022

Molecular-scale Integration of Multi-modal Sensing and Neuromorphic Computing with Organic Electrochemical Transistors

Abstract: Bionic learning with fused sensing, memory and processing functions outperforms artificial neural networks running on silicon chips in terms of efficiency and footprint. However, digital hardware implementation of bionic learning suffers from device heterogeneity in sensors and processing cores, which incurs large hardware, energy and time overheads. Here, we present a universal solution to simultaneously perform multi-modal sensing, memory and processing using organic electrochemical transistors with designed architecture and tailored channel morphology, selective ion injection into the crystalline/amorphous regions. The resultant device work as either a volatile receptor that shows multi-modal sensing, or a non-volatile synapse that features record-high 10-bit analog states, low switching stochasticity and good retention without the integration of any extra devices. Homogeneous integration of such devices enables bionic learning functions such as conditioned reflex and real-time cardiac disease diagnose via reservoir computing, illustrating the promise for future smart edge health informatics.

cs.ET↗

arXiv2022

Phylogenetic Study of 2019-nCoV by Using Alignment Free Method (Evolutionary Bifurcation of Novel Coronavirus Mutants)

The phylogenetic tree of SARS-CoV-2 (nCov-19) viruses is reconstructed according to the similarity of genome sequences. The tree topology of Betacoronavirus is remarkably consistent with biologist's systematics. Because the tree construction contains enough information about virus mutants, it is suitable to study the evolutionary relationship between novel coronavirus mutants transmitted among humans. The emergences of 14 kinds of main mutants are studied and these strains can be classified as eight bifurcations of the phylogenetic tree. It is found that there exist three types of virus mutations, namely, the mutation among sub-branches of the same branch, the off-root mutation and the root-oriented mutation between large branches of the tree. From the point of the relation between viral mutation and host selection we found that individuals with low immunity provide a special environment for the positive natural selection of virus evolution. It gives a mechanism to explain why large mutations between two distant branches generally occur in the nCov-19 phylogenetic tree. The finding is helpful to formulate strategies to control the spread of COVID-19.

q-bio.GN↗

arXiv2021

Multi-speaker Multi-style Text-to-speech Synthesis With Single-speaker Single-style Training Data Scenarios

In the existing cross-speaker style transfer task, a source speaker with multi-style recordings is necessary to provide the style for a target speaker. However, it is hard for one speaker to express all expected styles. In this paper, a more general task, which is to produce expressive speech by combining any styles and timbres from a multi-speaker corpus in which each speaker has a unique style, is proposed. To realize this task, a novel method is proposed. This method is a Tacotron2-based framework but with a fine-grained text-based prosody predicting module and a speaker identity controller. Experiments demonstrate that the proposed method can successfully express a style of one speaker with the timber of another speaker bypassing the dependency on a single speaker's multi-style corpus. Moreover, the explicit prosody features used in the prosody predicting module can increase the diversity of synthetic speech by adjusting the value of prosody features.

eess.AS↗

arXiv2021

M2DGR: A Multi-sensor and Multi-scenario SLAM Dataset for Ground Robots

We introduce M2DGR: a novel large-scale dataset collected by a ground robot with a full sensor-suite including six fish-eye and one sky-pointing RGB cameras, an infrared camera, an event camera, a Visual-Inertial Sensor (VI-sensor), an inertial measurement unit (IMU), a LiDAR, a consumer-grade Global Navigation Satellite System (GNSS) receiver and a GNSS-IMU navigation system with real-time kinematic (RTK) signals. All those sensors were well-calibrated and synchronized, and their data were recorded simultaneously. The ground truth trajectories were obtained by the motion capture device, a laser 3D tracker, and an RTK receiver. The dataset comprises 36 sequences (about 1TB) captured in diverse scenarios including both indoor and outdoor environments. We evaluate state-of-the-art SLAM algorithms on M2DGR. Results show that existing solutions perform poorly in some scenarios. For the benefit of the research community, we make the dataset and tools public. The webpage of our project is https://github.com/SJTU-ViSYS/M2DGR.

cs.RO↗

arXiv2021

Elastic Significant Bit Quantization and Acceleration for Deep Neural Networks

Quantization has been proven to be a vital method for improving the inference efficiency of deep neural networks (DNNs). However, it is still challenging to strike a good balance between accuracy and efficiency while quantizing DNN weights or activation values from high-precision formats to their quantized counterparts. We propose a new method called elastic significant bit quantization (ESB) that controls the number of significant bits of quantized values to obtain better inference accuracy with fewer resources. We design a unified mathematical formula to constrain the quantized values of the ESB with a flexible number of significant bits. We also introduce a distribution difference aligner (DDA) to quantitatively align the distributions between the full-precision weight or activation values and quantized values. Consequently, ESB is suitable for various bell-shaped distributions of weights and activation of DNNs, thus maintaining a high inference accuracy. Benefitting from fewer significant bits of quantized values, ESB can reduce the multiplication complexity. We implement ESB as an accelerator and quantitatively evaluate its efficiency on FPGAs. Extensive experimental results illustrate that ESB quantization consistently outperforms state-of-the-art methods and achieves average accuracy improvements of 4.78%, 1.92%, and 3.56% over AlexNet, ResNet18, and MobileNetV2, respectively. Furthermore, ESB as an accelerator can achieve 10.95 GOPS peak performance of 1k LUTs without DSPs on the Xilinx ZCU102 FPGA platform. Compared with CPU, GPU, and state-of-the-art accelerators on FPGAs, the ESB accelerator can improve the energy efficiency by up to 65x, 11x, and 26x, respectively.

cs.CV↗

arXiv2021

A route to engineered high aspect-ratio silicon nanostructures through regenerative secondary mask lithography

Silicon nanostructuring imparts unique material properties including antireflectivity, antifogging, anti-icing, self-cleaning, and/or antimicrobial activity. To tune these properties however, a good control over features size and shape is essential. Here, a versatile fabrication process is presented to achieve tailored silicon nanostructures (thin/thick pillars, sharp/truncated/re-entrant cones), of pitch down to ~50 nm, and high-aspect ratio (>10). The approach relies on pre-assembled block copolymer (BCP) micelles and their direct transfer into a glass hard mask of an arbitrary thickness, now enabled by our recently reported regenerative secondary mask lithography. During this pattern transfer, not only the mask diameter can be decreased but also uniquely increased; constituting the first method to achieve such tunability without necessitating a different molecular weight BCP. Consequently, the hard mask modulation (height, diameter) advances the flexibility in attainable inter-pillar spacing, aspect ratios, and re-entrant profiles (= glass on silicon). Combined with adjusted silicon etch conditions, the morphology of nanopatterns can be highly customized. The process control and scalability enable uniform patterning of a 6-inch wafer which is verified through cross-wafer excellent antireflectivity (<5%) and water-repellency (advancing contact angle 158°; hysteresis 1°). It is envisioned the implementation of this approach to silicon nanostructuring to be far-reaching, facilitating fundamental studies and targeting applications spanning solar panels, antifogging/antibacterial surfaces, sensing, amongst many others.

physics.chem-ph↗