arXiv subjects

Yi Li

Publications and source records attributed to Yi Li.

At least 415 records · Page 23Linked to original sources

arXiv2022

WSSS4LUAD: Grand Challenge on Weakly-supervised Tissue Semantic Segmentation for Lung Adenocarcinoma

Lung cancer is the leading cause of cancer death worldwide, and adenocarcinoma (LUAD) is the most common subtype. Exploiting the potential value of the histopathology images can promote precision medicine in oncology. Tissue segmentation is the basic upstream task of histopathology image analysis. Existing deep learning models have achieved superior segmentation performance but require sufficient pixel-level annotations, which is time-consuming and expensive. To enrich the label resources of LUAD and to alleviate the annotation efforts, we organize this challenge WSSS4LUAD to call for the outstanding weakly-supervised semantic segmentation (WSSS) techniques for histopathology images of LUAD. Participants have to design the algorithm to segment tumor epithelial, tumor-associated stroma and normal tissue with only patch-level labels. This challenge includes 10,091 patch-level annotations (the training set) and over 130 million labeled pixels (the validation and test sets), from 87 WSIs (67 from GDPH, 20 from TCGA). All the labels were generated by a pathologist-in-the-loop pipeline with the help of AI models and checked by the label review board. Among 532 registrations, 28 teams submitted the results in the test phase with over 1,000 submissions. Finally, the first place team achieved mIoU of 0.8413 (tumor: 0.8389, stroma: 0.7931, normal: 0.8919). According to the technical reports of the top-tier teams, CAM is still the most popular approach in WSSS. Cutmix data augmentation has been widely adopted to generate more reliable samples. With the success of this challenge, we believe that WSSS approaches with patch-level annotations can be a complement to the traditional pixel annotations while reducing the annotation efforts. The entire dataset has been released to encourage more researches on computational pathology in LUAD and more novel WSSS techniques.

eess.IV↗

arXiv2022

Tunable magnetically induced transparency spectra in magnon-magnon coupled Y3Fe5O12/permalloy bilayers

Hybrid magnonic systems host a variety of characteristic quantum phenomena such as the magnetically-induced transparency (MIT) and Purcell effect, which are considered useful for future coherent quantum information processing. In this work, we experimentally demonstrate a tunable MIT effect in the Y3Fe5O12(YIG)/Permalloy(Py) magnon-magnon coupled system via changing the magnetic field orientations. By probing the magneto-optic effects of Py and YIG, we identify clear features of MIT spectra induced by the mode hybridization between the uniform mode of Py and the perpendicular standing spin-wave modes of YIG. By changing the external magnetic field orientations, we observe a tunable coupling strength between the YIG's spin-wave modes and the Py's uniform mode, upon the application of an out-of-plane magnetic field. This observation is theoretically interpreted by a geometrical consideration of the Py and YIG magnetization under the oblique magnetic field even at a constant interfacial exchange coupling. Our findings show high promise for investigating tunable coherent phenomena with hybrid magnonic platforms.

cond-mat.mtrl-sci↗

arXiv2022

Hybrid magnonics for short-wavelength spin waves facilitated by a magnetic heterostructure

Recent research on hybrid magnonics has been restricted by the long magnon wavelengths of the ferromagnetic resonance modes. We present an experiment on the hybridization of 250-nm wavelength magnons with microwave photons in a multimode magnonic system consists of a planar cavity and a magnetic bilayer. The coupling between magnon modes in the two magnetic layers, i.e., the uniform mode in Permalloy (Py) and the perpendicular standing spin waves (PSSWs) in YIG, serves as an effective means for exciting short-wavelength PSSWs, which is further hybridized with the photon mode of the microwave resonator. The demonstrated magnon-photon coupling approaches the superstrong coupling regime, and can even be achieved near zero bias field.

cond-mat.mes-hall↗

arXiv2022

Unsupervised Image Deraining: Optimization Model Driven Deep CNN

The deep convolutional neural network has achieved significant progress for single image rain streak removal. However, most of the data-driven learning methods are full-supervised or semi-supervised, unexpectedly suffering from significant performance drops when dealing with real rain. These data-driven learning methods are representative yet generalize poor for real rain. The opposite holds true for the model-driven unsupervised optimization methods. To overcome these problems, we propose a unified unsupervised learning framework which inherits the generalization and representation merits for real rain removal. Specifically, we first discover a simple yet important domain knowledge that directional rain streak is anisotropic while the natural clean image is isotropic, and formulate the structural discrepancy into the energy function of the optimization model. Consequently, we design an optimization model-driven deep CNN in which the unsupervised loss function of the optimization model is enforced on the proposed network for better generalization. In addition, the architecture of the network mimics the main role of the optimization models with better feature representation. On one hand, we take advantage of the deep network to improve the representation. On the other hand, we utilize the unsupervised loss of the optimization model for better generalization. Overall, the unsupervised learning framework achieves good generalization and representation: unsupervised training (loss) with only a few real rainy images (input) and physical meaning network (architecture). Extensive experiments on synthetic and real-world rain datasets show the superiority of the proposed method.

cs.CV↗

arXiv2022

PEAR: Personalized Re-ranking with Contextualized Transformer for Recommendation

The goal of recommender systems is to provide ordered item lists to users that best match their interests. As a critical task in the recommendation pipeline, re-ranking has received increasing attention in recent years. In contrast to conventional ranking models that score each item individually, re-ranking aims to explicitly model the mutual influences among items to further refine the ordering of items given an initial ranking list. In this paper, we present a personalized re-ranking model (dubbed PEAR) based on contextualized transformer. PEAR makes several major improvements over the existing methods. Specifically, PEAR not only captures feature-level and item-level interactions, but also models item contexts from both the initial ranking list and the historical clicked item list. In addition to item-level ranking score prediction, we also augment the training of PEAR with a list-level classification task to assess users' satisfaction on the whole ranking list. Experimental results on both public and production datasets have shown the superior effectiveness of PEAR compared to the previous re-ranking models.

cs.IR↗

arXiv2022

Geometric Effect of High-Resolution Electron Energy Loss Spectroscopy on the Identification of Plasmons: An Example of Graphene

High-resolution electron energy loss spectroscopy (HREELS) is one of the most powerful methods to detect the dispersion of plasmons. However, we find that in the HREELS measurement, the scattering geometric configuration will seriously affect the identification of plasmons. Here, taking graphene as an example, using the HREELS capable of two-dimensional energy-momentum mapping combined with the intensity distribution calculations, we visually display the intensity distribution of the scattering geometric factor. We demonstrate that the energy loss peaks from the scattering geometric effect may be misinterpreted as the features of an acoustic plasmon. In any HREELS measurement, it is necessary to evaluate the effect of the scattering geometry quantitatively to identify the intrinsic surface excitations.

cond-mat.mes-hall↗

arXiv2022

Near unity Raman $β$-factor of surface enhanced Raman scattering in a waveguide

The Raman scattering of light by molecular vibrations offers a powerful technique to 'fingerprint' molecules via their internal bonds and symmetries. Since Raman scattering is weak, methods to enhance, direct and harness it are highly desirable, e.g. through the use of optical cavities, waveguides, and surface enhanced Raman scattering (SERS). While SERS offers dramatic enhancements by localizing light within vanishingly small 'hot-spots' in metallic nanostructures, these tiny interaction volumes are only sensitive to few molecules, yielding weak signals that are difficult to detect. Here, we show that SERS from 4-Aminothiophenol (4-ATP) molecules bonded to a plasmonic gap waveguide is directed into a single mode with >99% efficiency. Although sacrificing a confinement dimension, we find 10$^4$ times SERS enhancement across a broad spectral range enabled by the waveguide's larger sensing volume and non-resonant mode. Remarkably, the waveguide-SERS (W-SERS) is bright enough to image Raman transport across the waveguides exposing the roles of nanofocusing and the Purcell effect. Emulating the $β$-factor from laser physics, the near unity Raman $β$-factor observed exposes the SERS technique in a new light and points to alternative routes to controlling Raman scattering. The ability of W-SERS to direct Raman scattering is relevant to Raman sensors based on integrated photonics with applications in gas and bio-sensing as well as healthcare.

physics.optics↗

arXiv2022

Tight Bounds for Sketching the Operator Norm, Schatten Norms, and Subspace Embeddings

We consider the following oblivious sketching problem: given $ε\in (0,1/3)$ and $n \geq d/ε^2$, design a distribution $\mathcal{D}$ over $\mathbb{R}^{k \times nd}$ and a function $f: \mathbb{R}^k \times \mathbb{R}^{nd} \rightarrow \mathbb{R}$, so that for any $n \times d$ matrix $A$, $$\Pr_{S \sim \mathcal{D}} [(1-ε) \|A\|_{op} \leq f(S(A),S) \leq (1+ε)\|A\|_{op}] \geq 2/3,$$ where $\|A\|_{op}$ is the operator norm of $A$ and $S(A)$ denotes $S \cdot A$, interpreting $A$ as a vector in $\mathbb{R}^{nd}$. We show a tight lower bound of $k = Ω(d^2/ε^2)$ for this problem. Our result considerably strengthens the result of Nelson and Nguyen (ICALP, 2014), as it (1) applies only to estimating the operator norm, which can be estimated given any OSE, and (2) applies to distributions over general linear operators $S$ which treat $A$ as a vector and compute $S(A)$, rather than the restricted class of linear operators corresponding to matrix multiplication. Our technique also implies the first tight bounds for approximating the Schatten $p$-norm for even integers $p$ via general linear sketches, improving the previous lower bound from $k = Ω(n^{2-6/p})$ [Regev, 2014] to $k = Ω(n^{2-4/p})$. Importantly, for sketching the operator norm up to a factor of $α$, where $α- 1 = Ω(1)$, we obtain a tight $k = Ω(n^2/α^4)$ bound, matching the upper bound of Andoni and Nguyen (SODA, 2013), and improving the previous $k = Ω(n^2/α^6)$ lower bound. Finally, we also obtain the first lower bounds for approximating Ky Fan norms.

cs.DS↗

arXiv2022

High Dimensional Gaussian Graphical Regression Models with Covariates

Though Gaussian graphical models have been widely used in many scientific fields, relatively limited progress has been made to link graph structures to external covariates. We propose a Gaussian graphical regression model, which regresses both the mean and the precision matrix of a Gaussian graphical model on covariates. In the context of co-expression quantitative trait locus (QTL) studies, our method can determine how genetic variants and clinical conditions modulate the subject-level network structures, and recover both the population-level and subject-level gene networks. Our framework encourages sparsity of covariate effects on both the mean and the precision matrix. In particular for the precision matrix, we stipulate simultaneous sparsity, i.e., group sparsity and element-wise sparsity, on effective covariates and their effects on network edges, respectively. We establish variable selection consistency first under the case with known mean parameters and then a more challenging case with unknown means depending on external covariates, and establish in both cases the $\ell_2$ convergence rates and the selection consistency of the estimated precision parameters. The utility and efficacy of our proposed method is demonstrated through simulation studies and an application to a co-expression QTL study with brain cancer patients.

stat.ME↗

arXiv2022

Global and Local Contrastive Self-Supervised Learning for Semantic Segmentation of HR Remote Sensing Images

Supervised learning for semantic segmentation requires a large number of labeled samples, which is difficult to obtain in the field of remote sensing. Self-supervised learning (SSL), can be used to solve such problems by pre-training a general model with a large number of unlabeled images and then fine-tuning it on a downstream task with very few labeled samples. Contrastive learning is a typical method of SSL that can learn general invariant features. However, most existing contrastive learning methods are designed for classification tasks to obtain an image-level representation, which may be suboptimal for semantic segmentation tasks requiring pixel-level discrimination. Therefore, we propose a global style and local matching contrastive learning network (GLCNet) for remote sensing image semantic segmentation. Specifically, 1) the global style contrastive learning module is used to better learn an image-level representation, as we consider that style features can better represent the overall image features. 2) The local features matching contrastive learning module is designed to learn representations of local regions, which is beneficial for semantic segmentation. The experimental results show that our method mostly outperforms SOTA self-supervised methods and the ImageNet pre-training method. Specifically, with 1\% annotation from the original dataset, our approach improves Kappa by 6\% on the ISPRS Potsdam dataset relative to the existing baseline. Moreover, our method outperforms supervised learning methods when there are some differences between the datasets of upstream tasks and downstream tasks. Since SSL could directly learn the essential characteristics of data from unlabeled data, which is easy to obtain in the remote sensing field, this may be of great significance for tasks such as global mapping. The source code is available at https://github.com/GeoX-Lab/G-RSIM.

cs.CV↗

arXiv2022

Coherent coupling of two remote magnonic resonators mediated by superconducting circuits

We demonstrate microwave-mediated distant magnon-magnon coupling on a superconducting circuit platform, incorporating chip-mounted single-crystal Y$_3$Fe$_5$O$_{12}$ (YIG) spheres. Coherent level repulsion and dissipative level attraction between the magnon modes of the two YIG spheres are demonstrated. The former is mediated by cavity photons of a superconducting resonator, and the latter is mediated by propagating photons of a coplanar waveguide. Our results open new avenues towards exploring integrated hybrid magnonic networks for coherent information processing on a quantum-compatible superconducting platform.

cond-mat.mes-hall↗

arXiv2021

Echo state graph neural networks with analogue random resistor arrays

Recent years have witnessed an unprecedented surge of interest, from social networks to drug discovery, in learning representations of graph-structured data. However, graph neural networks, the machine learning models for handling graph-structured data, face significant challenges when running on conventional digital hardware, including von Neumann bottleneck incurred by physically separated memory and processing units, slowdown of Moore's law due to transistor scaling limit, and expensive training cost. Here we present a novel hardware-software co-design, the random resistor array-based echo state graph neural network, which addresses these challenges. The random resistor arrays not only harness low-cost, nanoscale and stackable resistors for highly efficient in-memory computing using simple physical laws, but also leverage the intrinsic stochasticity of dielectric breakdown to implement random projections in hardware for an echo state network that effectively minimizes the training cost thanks to its fixed and random weights. The system demonstrates state-of-the-art performance on both graph classification using the MUTAG and COLLAB datasets and node classification using the CORA dataset, achieving 34.2x, 93.2x, and 570.4x improvement of energy efficiency and 98.27%, 99.46%, and 95.12% reduction of training cost compared to conventional graph learning on digital hardware, respectively, which may pave the way for the next generation AI system for graph learning.

cs.ET↗

arXiv2021

Lower Bounds for Sparse Oblivious Subspace Embeddings

An oblivious subspace embedding (OSE), characterized by parameters $m,n,d,ε,δ$, is a random matrix $Π\in \mathbb{R}^{m\times n}$ such that for any $d$-dimensional subspace $T\subseteq \mathbb{R}^n$, $\Pr_Π[\forall x\in T, (1-ε)\|x\|_2 \leq \|Πx\|_2\leq (1+ε)\|x\|_2] \geq 1-δ$. For $ε$ and $δ$ at most a small constant, we show that any OSE with one nonzero entry in each column must satisfy that $m = Ω(d^2/(ε^2δ))$, establishing the optimality of the classical Count-Sketch matrix. When an OSE has $1/(9ε)$ nonzero entries in each column, we show it must hold that $m = Ω(ε^{O(δ)} d^2)$, improving on the previous $Ω(ε^2 d^2)$ lower bound due to Nelson and Nguyen (ICALP 2014).

cs.DS↗

arXiv2021

Self-Supervised Learning based Monaural Speech Enhancement with Complex-Cycle-Consistent

Recently, self-supervised learning (SSL) techniques have been introduced to solve the monaural speech enhancement problem. Due to the lack of using clean phase information, the enhancement performance is limited in most SSL methods. Therefore, in this paper, we propose a phase-aware self-supervised learning based monaural speech enhancement method. The latent representations of both amplitude and phase are studied in two decoders of the foundation autoencoder (FAE) with only a limited set of clean speech signals independently. Then, the downstream autoencoder (DAE) learns a shared latent space between the clean speech and mixture representations with a large number of unseen mixtures. A complex-cycle-consistent (CCC) mechanism is proposed to minimize the reconstruction loss between the amplitude and phase domains. Besides, it is noticed that if the speech features are extracted as the multi-resolution spectra, the desired information distributed in spectra of different scales can be studied to further boost the performance. The NOISEX and DAPS corpora are used to generate mixtures with different interferences to evaluate the efficacy of the proposed method. It is highlighted that the clean speech and mixtures fed in FAE and DAE are not paired. Both ablation and comparison experimental results show that the proposed method clearly outperforms the state-of-the-art approaches.

cs.SD↗

arXiv2021

Self-Supervised Learning based Monaural Speech Enhancement with Multi-Task Pre-Training

In self-supervised learning, it is challenging to reduce the gap between the enhancement performance on the estimated and target speech signals with existed pre-tasks. In this paper, we propose a multi-task pre-training method to improve the speech enhancement performance with self-supervised learning. Within the pre-training autoencoder (PAE), only a limited set of clean speech signals are required to learn their latent representations. Meanwhile, to solve the limitation of single pre-task, the proposed masking module exploits the dereverberated mask and estimated ratio mask to denoise the mixture as the second pre-task. Different from the PAE, where the target speech signals are estimated, the downstream task autoencoder (DAE) utilizes a large number of unlabeled and unseen reverberant mixtures to generate the estimated mixtures. The trained DAE is shared by the learned representations and masks. Experimental results on a benchmark dataset demonstrate that the proposed method outperforms the state-of-the-art approaches.

cs.SD↗

arXiv2021

Uncertainty Estimation via Response Scaling for Pseudo-mask Noise Mitigation in Weakly-supervised Semantic Segmentation

Weakly-Supervised Semantic Segmentation (WSSS) segments objects without a heavy burden of dense annotation. While as a price, generated pseudo-masks exist obvious noisy pixels, which result in sub-optimal segmentation models trained over these pseudo-masks. But rare studies notice or work on this problem, even these noisy pixels are inevitable after their improvements on pseudo-mask. So we try to improve WSSS in the aspect of noise mitigation. And we observe that many noisy pixels are of high confidence, especially when the response range is too wide or narrow, presenting an uncertain status. Thus, in this paper, we simulate noisy variations of response by scaling the prediction map multiple times for uncertainty estimation. The uncertainty is then used to weight the segmentation loss to mitigate noisy supervision signals. We call this method URN, abbreviated from Uncertainty estimation via Response scaling for Noise mitigation. Experiments validate the benefits of URN, and our method achieves state-of-the-art results at 71.2% and 41.5% on PASCAL VOC 2012 and MS COCO 2014 respectively, without extra models like saliency detection. Code is available at https://github.com/XMed-Lab/URN.

cs.CV↗

arXiv2021

U-shaped Transformer with Frequency-Band Aware Attention for Speech Enhancement

The state-of-the-art speech enhancement has limited performance in speech estimation accuracy. Recently, in deep learning, the Transformer shows the potential to exploit the long-range dependency in speech by self-attention. Therefore, it is introduced in speech enhancement to improve the speech estimation accuracy from a noise mixture. However, to address the computational cost issue in Transformer with self-attention, the axial attention is the option i.e., to split a 2D attention into two 1D attentions. Inspired by the axial attention, in the proposed method we calculate the attention map along both time- and frequency-axis to generate time and frequency sub-attention maps. Moreover, different from the axial attention, the proposed method provides two parallel multi-head attentions for time- and frequency-axis. Furthermore, it is proven in the literature that the lower frequency-band in speech, generally, contains more desired information than the higher frequency-band, in a noise mixture. Therefore, the frequency-band aware attention is proposed i.e., high frequency-band attention (HFA), and low frequency-band attention (LFA). The U-shaped Transformer is also first time introduced in the proposed method to further improve the speech estimation accuracy. The extensive evaluations over four public datasets, confirm the efficacy of the proposed method.

cs.SD↗

arXiv2021

A local curvature estimate for the Ricci-harmonic flow on complete Riemannian manifolds

In this paper we consider the local $L^p$ estimate of Riemannian curvature for the Ricci-harmonic flow or List's flow introduced by List \cite{List2005} on complete noncompact manifolds. As an application, under the assumption that the flow exists on a finite time interval $[0,T)$ and the Ricci curvature is uniformly bounded, we prove that the $L^p$ norm of Riemannian curvature is bounded, and then, applying the De Giorgi-Nash-Moser iteration method, obtain the local boundedness of Riemannian curvature and consequently the flow can be continuously extended past $T$.

math.DG↗