Searcharxiv⌕ Search

arXiv subjects

Sheng Liu

Publications and source records attributed to Sheng Liu.

At least 109 records · Page 6Linked to original sources
arXiv2023

Unleashing the Potential of Regularization Strategies in Learning with Noisy Labels

In recent years, research on learning with noisy labels has focused on devising novel algorithms that can achieve robustness to noisy training labels while generalizing to clean data. These algorithms often incorporate sophisticated techniques, such as noise modeling, label correction, and co-training. In this study, we demonstrate that a simple baseline using cross-entropy loss, combined with widely used regularization strategies like learning rate decay, model weights average, and data augmentations, can outperform state-of-the-art methods. Our findings suggest that employing a combination of regularization strategies can be more effective than intricate algorithms in tackling the challenges of learning with noisy labels. While some of these regularization strategies have been utilized in previous noisy label learning research, their full potential has not been thoroughly explored. Our results encourage a reevaluation of benchmarks for learning with noisy labels and prompt reconsideration of the role of specialized learning algorithms designed for training with noisy labels.

cs.LG↗
arXiv2023

Imaging emergent exotic quasiparticle state in a frustrated transition metal oxide

The existence of rich Fermiology in anomalous metal phase in exotic superconductors has attracted considerable interests, as exemplified in copper, iron-based, and intermetallic frustrated kagome-based compounds. A common feature in these cases is pseudo-gap opening or long-range lattice/electronic ordering above superconducting critical temperature Tc. As yet developed area is the potential existence of exotic Fermiology in superconducting transition metal oxides on a geometrically frustrated lattice. Here, we focus on the spinel oxide superconductor LiTi2O4, which can be viewed as the hole-doped side of the orbital ordered 3d1 Mott system on the Ti-derived pyrochlore frustrated network. By the in-situ combination of angle-resolved photoemission spectroscopy (ARPES) and epitaxial thin film growth, we discovered the abrupt flattening of near Fermi energy dispersion below the characteristic temperature T* ~ 150 K. While the emergent negative thermal expansion below T* strongly supports a distinct phase at low-temperature, absence of energy gap opening, splitting/folding of bands, nor long-range lattice distortion are seen across T*. We propose that the competition between growing instability towards orbital ordering and its inherent geometric frustration in the Ti-pyrochlore network results in a new quantum state of matter with robust high entropic nature below T*. Our findings collectively point to a unique Fermiology in frustrated three-dimensional transition metal oxides, and its connection to superconductivity below Tc is open as an interesting future challenge. Also, a potential guideline is unexpectedly provided for designing zero thermal expansion metal to develop future solid-state devices.

cond-mat.supr-con↗
arXiv2023

Energy-time Entanglement Coexisting with Fiber Optical Communication at Telecom C-band

The coexistence of quantum and classical light in the same fiber link is extremely desired in developing quantum communication. It has been implemented for different quantum information tasks, such as classical light coexisting with polarization-entangled photons at telecom O-band, and with quantum signal based quantum key distribution (QKD). In this work, we demonstrate the coexistence of energy-time entanglement based QKD and fiber optical communication at the telecom C-band. The property of noise from the classical channel is characterized with classical light at different wavelengths. With the largest noise, i.e., the worst case, the properties of energy-time entanglement are measured at different fiber optical communication rates. By measuring the two-photon interference of energy-time entanglement, our results show that a visibility of 82.01$\pm$1.10\% is achieved with a bidirectional 20 Gbps fiber optical communication over 40 km. Furthermore, by performing the BBM92 protocol for QKD, a secret key rate of 245 bits per second could be generated with a quantum bit error rate of 8.88\% with the coexisted energy-time entanglement.~Our demonstration paves the way for developing the infrastructure for quantum networks compatible with fiber optical communication.

quant-ph↗
arXiv2023

Recovering the Block-wise Relationship in an Encryption-Then-Compression System

Joint encryption and compression is an ideal solution for protecting security and privacy of image data in a real scenario, e.g. storing them on an existing cloud-based service like Facebook. Recently, some block-wise encryption-then-compression (ETC) schemes compatible with JPEG were proposed to provide a reasonably high level of security without compromising compression ratio much. This paper investigates recovering the block-wise relationship in an ETC scheme exerting on single-color blocks of size $8\times 8$ in the scenarios of ciphertext-only attack, known-plaintext attack and chosen-plaintext attack. Then, the attacking targets are extended to the other conventional ETC schemes exerting on multiple color channels and blocks of various sizes. Especially, an elaborate jigsaw puzzle solver is designed to recover enough visual information from multiple cipher-images encrypted by the same secret key. Moreover, the nice attacking performance was verified over two social media platforms, Facebook and Weibo.

cs.CR↗
arXiv2023

LEMaRT: Label-Efficient Masked Region Transform for Image Harmonization

We present a simple yet effective self-supervised pre-training method for image harmonization which can leverage large-scale unannotated image datasets. To achieve this goal, we first generate pre-training data online with our Label-Efficient Masked Region Transform (LEMaRT) pipeline. Given an image, LEMaRT generates a foreground mask and then applies a set of transformations to perturb various visual attributes, e.g., defocus blur, contrast, saturation, of the region specified by the generated mask. We then pre-train image harmonization models by recovering the original image from the perturbed image. Secondly, we introduce an image harmonization model, namely SwinIH, by retrofitting the Swin Transformer [27] with a combination of local and global self-attention mechanisms. Pre-training SwinIH with LEMaRT results in a new state of the art for image harmonization, while being label-efficient, i.e., consuming less annotated data for fine-tuning than existing methods. Notably, on iHarmony4 dataset [8], SwinIH outperforms the state of the art, i.e., SCS-Co [16] by a margin of 0.4 dB when it is fine-tuned on only 50% of the training data, and by 1.0 dB when it is trained on the full training dataset.

cs.CV↗
arXiv2023

Avoiding spurious correlations via logit correction

Empirical studies suggest that machine learning models trained with empirical risk minimization (ERM) often rely on attributes that may be spuriously correlated with the class labels. Such models typically lead to poor performance during inference for data lacking such correlations. In this work, we explicitly consider a situation where potential spurious correlations are present in the majority of training data. In contrast with existing approaches, which use the ERM model outputs to detect the samples without spurious correlations and either heuristically upweight or upsample those samples, we propose the logit correction (LC) loss, a simple yet effective improvement on the softmax cross-entropy loss, to correct the sample logit. We demonstrate that minimizing the LC loss is equivalent to maximizing the group-balanced accuracy, so the proposed LC could mitigate the negative impacts of spurious correlations. Our extensive experimental results further reveal that the proposed LC loss outperforms state-of-the-art solutions on multiple popular benchmarks by a large margin, an average 5.5\% absolute improvement, without access to spurious attribute labels. LC is also competitive with oracle methods that make use of the attribute labels. Code is available at https://github.com/shengliu66/LC.

cs.LG↗
arXiv2023

Swin MAE: Masked Autoencoders for Small Datasets

The development of deep learning models in medical image analysis is majorly limited by the lack of large-sized and well-annotated datasets. Unsupervised learning does not require labels and is more suitable for solving medical image analysis problems. However, most of the current unsupervised learning methods need to be applied to large datasets. To make unsupervised learning applicable to small datasets, we proposed Swin MAE, which is a masked autoencoder with Swin Transformer as its backbone. Even on a dataset of only a few thousand medical images and without using any pre-trained models, Swin MAE is still able to learn useful semantic features purely from images. It can equal or even slightly outperform the supervised model obtained by Swin Transformer trained on ImageNet in terms of the transfer learning results of downstream tasks. The code is publicly available at https://github.com/Zian-Xu/Swin-MAE.

cs.CV↗
arXiv2022

Parotid Gland MRI Segmentation Based on Swin-Unet and Multimodal Images

Background and objective: Parotid gland tumors account for approximately 2% to 10% of head and neck tumors. Preoperative tumor localization, differential diagnosis, and subsequent selection of appropriate treatment for parotid gland tumors are critical. However, the relative rarity of these tumors and the highly dispersed tissue types have left an unmet need for a subtle differential diagnosis of such neoplastic lesions based on preoperative radiomics. Recently, deep learning methods have developed rapidly, especially Transformer beats the traditional convolutional neural network in computer vision. Many new Transformer-based networks have been proposed for computer vision tasks. Methods: In this study, multicenter multimodal parotid gland MR images were collected. The Swin-Unet which was based on Transformer was used. MR images of short time inversion recovery, T1-weighted and T2-weighted modalities were combined into three-channel data to train the network. We achieved segmentation of the region of interest for parotid gland and tumor. Results: The Dice-Similarity Coefficient of the model on the test set was 88.63%, Mean Pixel Accuracy was 99.31%, Mean Intersection over Union was 83.99%, and Hausdorff Distance was 3.04. Then a series of comparison experiments were designed in this paper to further validate the segmentation performance of the algorithm. Conclusions: Experimental results showed that our method has good results for parotid gland and tumor segmentation. The Transformer-based network outperforms the traditional convolutional neural network in the field of medical images.

eess.IV↗
arXiv2022

PADDLES: Phase-Amplitude Spectrum Disentangled Early Stopping for Learning with Noisy Labels

Convolutional Neural Networks (CNNs) have demonstrated superiority in learning patterns, but are sensitive to label noises and may overfit noisy labels during training. The early stopping strategy averts updating CNNs during the early training phase and is widely employed in the presence of noisy labels. Motivated by biological findings that the amplitude spectrum (AS) and phase spectrum (PS) in the frequency domain play different roles in the animal's vision system, we observe that PS, which captures more semantic information, can increase the robustness of DNNs to label noise, more so than AS can. We thus propose early stops at different times for AS and PS by disentangling the features of some layer(s) into AS and PS using Discrete Fourier Transform (DFT) during training. Our proposed Phase-AmplituDe DisentangLed Early Stopping (PADDLES) method is shown to be effective on both synthetic and real-world label-noise datasets. PADDLES outperforms other early stopping methods and obtains state-of-the-art performance.

cs.CV↗
arXiv2022

On-Demand Delivery from Stores: Dynamic Dispatching and Routing with Random Demand

On-demand delivery has become increasingly popular around the world. Motivated by a large grocery chain store who offers fast on-demand delivery services, we model and solve a stochastic dynamic driver dispatching and routing problem for last-mile delivery systems where on-time performance is the main target. The system operator needs to dispatch a set of drivers and specify their delivery routes facing random demand that arrives over a fixed number of periods. The resulting stochastic dynamic program is challenging to solve due to the curse of dimensionality. We propose a novel structured approximation framework to approximate the value function via a parametrized dispatching and routing policy. We analyze the structural properties of the approximation framework and establish its performance guarantee under large-demand scenarios. We then develop efficient exact algorithms for the approximation problem based on Benders decomposition and column generation, which deliver verifiably optimal solutions within minutes. The evaluation results on a real-world data set show that our framework outperforms the current policy of the company by 36.53% on average in terms of delivery time. We also perform several policy experiments to understand the value of dynamic dispatching and routing with varying fleet sizes and dispatch frequencies.

math.OC↗
arXiv2022

Deep Probability Estimation

Reliable probability estimation is of crucial importance in many real-world applications where there is inherent (aleatoric) uncertainty. Probability-estimation models are trained on observed outcomes (e.g. whether it has rained or not, or whether a patient has died or not), because the ground-truth probabilities of the events of interest are typically unknown. The problem is therefore analogous to binary classification, with the difference that the objective is to estimate probabilities rather than predicting the specific outcome. This work investigates probability estimation from high-dimensional data using deep neural networks. There exist several methods to improve the probabilities generated by these models but they mostly focus on model (epistemic) uncertainty. For problems with inherent uncertainty, it is challenging to evaluate performance without access to ground-truth probabilities. To address this, we build a synthetic dataset to study and compare different computable metrics. We evaluate existing methods on the synthetic data as well as on three real-world probability estimation tasks, all of which involve inherent uncertainty: precipitation forecasting from radar images, predicting cancer patient survival from histopathology images, and predicting car crashes from dashcam videos. We also give a theoretical analysis of a model for high-dimensional probability estimation which reproduces several of the phenomena evinced in our experiments. Finally, we propose a new method for probability estimation using neural networks, which modifies the training process to promote output probabilities that are consistent with empirical probabilities computed from the data. The method outperforms existing approaches on most metrics on the simulated as well as real-world data.

cs.LG↗
arXiv2022

Are All Losses Created Equal: A Neural Collapse Perspective

While cross entropy (CE) is the most commonly used loss to train deep neural networks for classification tasks, many alternative losses have been developed to obtain better empirical performance. Among them, which one is the best to use is still a mystery, because there seem to be multiple factors affecting the answer, such as properties of the dataset, the choice of network architecture, and so on. This paper studies the choice of loss function by examining the last-layer features of deep networks, drawing inspiration from a recent line work showing that the global optimal solution of CE and mean-square-error (MSE) losses exhibits a Neural Collapse phenomenon. That is, for sufficiently large networks trained until convergence, (i) all features of the same class collapse to the corresponding class mean and (ii) the means associated with different classes are in a configuration where their pairwise distances are all equal and maximized. We extend such results and show through global solution and landscape analyses that a broad family of loss functions including commonly used label smoothing (LS) and focal loss (FL) exhibits Neural Collapse. Hence, all relevant losses(i.e., CE, LS, FL, MSE) produce equivalent features on training data. Based on the unconstrained feature model assumption, we provide either the global landscape analysis for LS loss or the local landscape analysis for FL loss and show that the (only!) global minimizers are neural collapse solutions, while all other critical points are strict saddles whose Hessian exhibit negative curvature directions either in the global scope for LS loss or in the local scope for FL loss near the optimal solution. The experiments further show that Neural Collapse features obtained from all relevant losses lead to largely identical performance on test data as well, provided that the network is sufficiently large and trained until convergence.

cs.LG↗
arXiv2022

Uncertainty Detection and Reduction in Neural Decoding of EEG Signals

EEG decoding systems based on deep neural networks have been widely used in decision making of brain computer interfaces (BCI). Their predictions, however, can be unreliable given the significant variance and noise in EEG signals. Previous works on EEG analysis mainly focus on the exploration of noise pattern in the source signal, while the uncertainty during the decoding process is largely unexplored. Automatically detecting and reducing such decoding uncertainty is important for BCI motor imagery applications such as robotic arm control etc. In this work, we proposed an uncertainty estimation and reduction model (UNCER) to quantify and mitigate the uncertainty during the EEG decoding process. It utilized a combination of dropout oriented method and Bayesian neural network for uncertainty estimation to incorporate both the uncertainty in the input signal and the uncertainty in the model parameters. We further proposed a data augmentation based approach for uncertainty reduction. The model can be integrated into current widely used EEG neural decoders without change of architecture. We performed extensive experiments for uncertainty estimation and its reduction in both intra-subject EEG decoding and cross-subject EEG decoding on two public motor imagery datasets, where the proposed model achieves significant improvement both on the quality of estimated uncertainty and the effectiveness of uncertainty reduction.

eess.SP↗
arXiv2022

Cryptanalyzing an Image Encryption Algorithm Underpinned by 2D Lag-Complex Logistic Map

This paper analyzes security performance of an image encryption algorithm using 2D lag-complex Logistic map (LCLM), which adopts it as a pseudo-random number generator, and uses the sum of all pixel values of the plain-image as its initial value to control the random combination of the basic encryption operations. However, multiple factors make the final pseudo-random sequences controlling the encryption process may be the same for different plain-images. Based on this point, we proposed a chosen-plaintext attack by attacking the six encryption steps with a strategy of divide and conquer. Using the pitfalls of 2D-LCLM, the number of required chosen plain-images is further reduced to $5\cdot\log_2(MN)+95$, where $\mathit{MN}$ is the number of pixels of the plain-image.

cs.CR↗
arXiv2022

Robust Training under Label Noise by Over-parameterization

Recently, over-parameterized deep networks, with increasingly more network parameters than training samples, have dominated the performances of modern machine learning. However, when the training data is corrupted, it has been well-known that over-parameterized networks tend to overfit and do not generalize. In this work, we propose a principled approach for robust training of over-parameterized deep networks in classification tasks where a proportion of training labels are corrupted. The main idea is yet very simple: label noise is sparse and incoherent with the network learned from clean data, so we model the noise and learn to separate it from the data. Specifically, we model the label noise via another sparse over-parameterization term, and exploit implicit algorithmic regularizations to recover and separate the underlying corruptions. Remarkably, when trained using such a simple method in practice, we demonstrate state-of-the-art test accuracy against label noise on a variety of real datasets. Furthermore, our experimental results are corroborated by theory on simplified linear models, showing that exact separation between sparse noise and low-rank data can be achieved under incoherent conditions. The work opens many interesting directions for improving over-parameterized models by using sparse over-parameterization and implicit regularization.

cs.LG↗
arXiv2022

On Learning Contrastive Representations for Learning with Noisy Labels

Deep neural networks are able to memorize noisy labels easily with a softmax cross-entropy (CE) loss. Previous studies attempted to address this issue focus on incorporating a noise-robust loss function to the CE loss. However, the memorization issue is alleviated but still remains due to the non-robust CE loss. To address this issue, we focus on learning robust contrastive representations of data on which the classifier is hard to memorize the label noise under the CE loss. We propose a novel contrastive regularization function to learn such representations over noisy data where label noise does not dominate the representation learning. By theoretically investigating the representations induced by the proposed regularization function, we reveal that the learned representations keep information related to true labels and discard information related to corrupted labels. Moreover, our theoretical results also indicate that the learned representations are robust to the label noise. The effectiveness of this method is demonstrated with experiments on benchmark datasets.

cs.LG↗
arXiv2022

Asymmetric Dual-Decoder U-Net for Joint Rain and Haze Removal

This work studies the joint rain and haze removal problem. In real-life scenarios, rain and haze, two often co-occurring common weather phenomena, can greatly degrade the clarity and quality of the scene images, leading to a performance drop in the visual applications, such as autonomous driving. However, jointly removing the rain and haze in scene images is ill-posed and challenging, where the existence of haze and rain and the change of atmosphere light, can both degrade the scene information. Current methods focus on the contamination removal part, thus ignoring the restoration of the scene information affected by the change of atmospheric light. We propose a novel deep neural network, named Asymmetric Dual-decoder U-Net (ADU-Net), to address the aforementioned challenge. The ADU-Net produces both the contamination residual and the scene residual to efficiently remove the rain and haze while preserving the fidelity of the scene information. Extensive experiments show our work outperforms the existing state-of-the-art methods by a considerable margin in both synthetic data and real-world data benchmarks, including RainCityscapes, BID Rain, and SPA-Data. For instance, we improve the state-of-the-art PSNR value by 2.26/4.57 on the RainCityscapes/SPA-Data, respectively. Codes will be made available freely to the research community.

cs.CV↗