PyTorchKR Papers

최신 논문

총 10,240편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2301.00981

Transfer Generative Adversarial Networks (T-GAN)-based Terahertz Channel Modeling

Terahertz (THz) communications are envisioned as a promising technology for 6G and beyond wireless systems, providing ultra-broad bandwidth and thus Terabit-per-second (Tbps) data rates. However, as foundation of designing THz communications, channel modeling and characterization are fundamental to scrutinize the potential of the new spectrum. Relied on physical measurements, traditional statistical channel modeling methods suffer from the problem of low accuracy with the assumed certain distributions and empirical parameters. Moreover, it is time-consuming and expensive to acquire extensive channel measurement in the THz band. In this paper, a transfer generative adversarial network (T-GAN) based modeling method is proposed in the THz band, which exploits the advantage of GAN in modeling the complex distribution, and the benefit of transfer learning in transferring the knowledge from a source task to improve generalization about the target task with limited training data. Specifically, to start with, the proposed GAN is pre-trained using the simulated dataset, generated by the standard channel model from 3rd generation partnerships project (3GPP). Furthermore, by transferring the knowledge and fine-tuning the pre-trained GAN, the T-GAN is developed by using the THz measured dataset with a small amount. Experimental results reveal that the distribution of PDPs generated by the proposed T-GAN method shows good agreement with measurement. Moreover, T-GAN achieves good performance in channel modeling, with 9 dB improved root-mean-square error (RMSE) and higher Structure Similarity Index Measure (SSIM), compared with traditional 3GPP method.

arXiv:2301.01233

Transferable Energy Storage Bidder

Energy storage resources must consider both price uncertainties and their physical operating characteristics when participating in wholesale electricity markets. This is a challenging problem as electricity prices are highly volatile, and energy storage has efficiency losses, power, and energy constraints. This paper presents a novel, versatile, and transferable approach combining model-based optimization with a convolutional long short-term memory network for energy storage to respond to or bid into wholesale electricity markets. We test our proposed approach using historical prices from New York State, showing it achieves state-of-the-art results, achieving between 70% to near 90% profit ratio compared to perfect foresight cases, in both price response and wholesale market bidding setting with various energy storage durations. We also test a transfer learning approach by pre-training the bidding model using New York data and applying it to arbitrage in Queensland, Australia. The result shows transfer learning achieves exceptional arbitrage profitability with as little as three days of local training data, demonstrating its significant advantage over training from scratch in scenarios with very limited data availability.

arXiv:2301.01162

Language Models are Drummers: Drum Composition with Natural Language Pre-Training

Automatic music generation with artificial intelligence typically requires a large amount of data which is hard to obtain for many less common genres and musical instruments. To tackle this issue, we present ongoing work and preliminary findings on the possibility for deep models to transfer knowledge from language to music, by finetuning large language models pre-trained on a massive text corpus on only hundreds of MIDI files of drum performances. We show that by doing so, one of the largest, state-of-the-art models (GPT3) is capable of generating reasonable drum grooves, while models that are not pre-trained (Transformer) shows no such ability beyond naive repetition. Evaluating generated music is a challenging task, more so is evaluating drum grooves with little precedence in literature. Hence, we propose a tailored structural evaluation method and analyze drum grooves produced by GPT3 compared to those played by human professionals, exposing the strengths and weaknesses of such generation by language-to-music transfer. Our findings suggest that language-to-music transfer learning with large language models is viable and promising.

arXiv:2301.01148

MERLIN: Multi-agent offline and transfer learning for occupant-centric energy flexible operation of grid-interactive communities using smart meter data and CityLearn

The decarbonization of buildings presents new challenges for the reliability of the electrical grid as a result of the intermittency of renewable energy sources and increase in grid load brought about by end-use electrification. To restore reliability, grid-interactive efficient buildings can provide flexibility services to the grid through demand response. Residential demand response programs are hindered by the need for manual intervention by customers. To maximize the energy flexibility potential of residential buildings, an advanced control architecture is needed. Reinforcement learning is well-suited for the control of flexible resources as it is able to adapt to unique building characteristics compared to expert systems. Yet, factors hindering the adoption of RL in real-world applications include its large data requirements for training, control security and generalizability. Here we address these challenges by proposing the MERLIN framework and using a digital twin of a real-world 17-building grid-interactive residential community in CityLearn. We show that 1) independent RL-controllers for batteries improve building and district level KPIs compared to a reference RBC by tailoring their policies to individual buildings, 2) despite unique occupant behaviours, transferring the RL policy of any one of the buildings to other buildings provides comparable performance while reducing the cost of training, 3) training RL-controllers on limited temporal data that does not capture full seasonality in occupant behaviour has little effect on performance. Although, the zero-net-energy (ZNE) condition of the buildings could be maintained or worsened as a result of controlled batteries, KPIs that are typically improved by ZNE condition (electricity price and carbon emissions) are further improved when the batteries are managed by an advanced controller.

arXiv:2301.01038

Heterogeneous Domain Adaptation and Equipment Matching: DANN-based Alignment with Cyclic Supervision (DBACS)

Process monitoring and control are essential in modern industries for ensuring high quality standards and optimizing production performance. These technologies have a long history of application in production and have had numerous positive impacts, but also hold great potential when integrated with Industry 4.0 and advanced machine learning, particularly deep learning, solutions. However, in order to implement these solutions in production and enable widespread adoption, the scalability and transferability of deep learning methods have become a focus of research. While transfer learning has proven successful in many cases, particularly with computer vision and homogenous data inputs, it can be challenging to apply to heterogeneous data. Motivated by the need to transfer and standardize established processes to different, non-identical environments and by the challenge of adapting to heterogeneous data representations, this work introduces the Domain Adaptation Neural Network with Cyclic Supervision (DBACS) approach. DBACS addresses the issue of model generalization through domain adaptation, specifically for heterogeneous data, and enables the transfer and scalability of deep learning-based statistical control methods in a general manner. Additionally, the cyclic interactions between the different parts of the model enable DBACS to not only adapt to the domains, but also match them. To the best of our knowledge, DBACS is the first deep learning approach to combine adaptation and matching for heterogeneous data settings. For comparison, this work also includes subspace alignment and a multi-view learning that deals with heterogeneous representations by mapping data into correlated latent feature spaces. Finally, DBACS with its ability to adapt and match, is applied to a virtual metrology use case for an etching process run on different machine types in semiconductor manufacturing.

arXiv:2301.00934

Finding the Most Transferable Tasks for Brain Image Segmentation

Although many studies have successfully applied transfer learning to medical image segmentation, very few of them have investigated the selection strategy when multiple source tasks are available for transfer. In this paper, we propose a prior knowledge guided and transferability based framework to select the best source tasks among a collection of brain image segmentation tasks, to improve the transfer learning performance on the given target task. The framework consists of modality analysis, RoI (region of interest) analysis, and transferability estimation, such that the source task selection can be refined step by step. Specifically, we adapt the state-of-the-art analytical transferability estimation metrics to medical image segmentation tasks and further show that their performance can be significantly boosted by filtering candidate source tasks based on modality and RoI characteristics. Our experiments on brain matter, brain tumor, and white matter hyperintensities segmentation datasets reveal that transferring from different tasks under the same modality is often more successful than transferring from the same task under different modalities. Furthermore, within the same modality, transferring from the source task that has stronger RoI shape similarity with the target task can significantly improve the final transfer performance. And such similarity can be captured using the Structural Similarity index in the label space.

arXiv:2301.00785공식 코드 · Papers with Code

CLIP-Driven Universal Model for Organ Segmentation and Tumor Detection

An increasing number of public datasets have shown a marked impact on automated organ segmentation and tumor detection. However, due to the small size and partially labeled problem of each dataset, as well as a limited investigation of diverse types of tumors, the resulting models are often limited to segmenting specific organs/tumors and ignore the semantics of anatomical structures, nor can they be extended to novel domains. To address these issues, we propose the CLIP-Driven Universal Model, which incorporates text embedding learned from Contrastive Language-Image Pre-training (CLIP) to segmentation models. This CLIP-based label encoding captures anatomical relationships, enabling the model to learn a structured feature embedding and segment 25 organs and 6 types of tumors. The proposed model is developed from an assembly of 14 datasets, using a total of 3,410 CT scans for training and then evaluated on 6,162 external CT scans from 3 additional datasets. We rank first on the Medical Segmentation Decathlon (MSD) public leaderboard and achieve state-of-the-art results on Beyond The Cranial Vault (BTCV). Additionally, the Universal Model is computationally more efficient (6x faster) compared with dataset-specific models, generalized better to CT scans from varying sites, and shows stronger transfer learning performance on novel tasks.

arXiv:2301.00265

Source-Free Unsupervised Domain Adaptation: A Survey

Unsupervised domain adaptation (UDA) via deep learning has attracted appealing attention for tackling domain-shift problems caused by distribution discrepancy across different domains. Existing UDA approaches highly depend on the accessibility of source domain data, which is usually limited in practical scenarios due to privacy protection, data storage and transmission cost, and computation burden. To tackle this issue, many source-free unsupervised domain adaptation (SFUDA) methods have been proposed recently, which perform knowledge transfer from a pre-trained source model to unlabeled target domain with source data inaccessible. A comprehensive review of these works on SFUDA is of great significance. In this paper, we provide a timely and systematic literature review of existing SFUDA approaches from a technical perspective. Specifically, we categorize current SFUDA studies into two groups, i.e., white-box SFUDA and black-box SFUDA, and further divide them into finer subcategories based on different learning strategies they use. We also investigate the challenges of methods in each subcategory, discuss the advantages/disadvantages of white-box and black-box SFUDA methods, conclude the commonly used benchmark datasets, and summarize the popular techniques for improved generalizability of models learned without using source data. We finally discuss several promising future directions in this field.

arXiv:2301.00355

Second Thoughts are Best: Learning to Re-Align With Human Values from Text Edits

We present Second Thought, a new learning paradigm that enables language models (LMs) to re-align with human values. By modeling the chain-of-edits between value-unaligned and value-aligned text, with LM fine-tuning and additional refinement through reinforcement learning, Second Thought not only achieves superior performance in three value alignment benchmark datasets but also shows strong human-value transfer learning ability in few-shot scenarios. The generated editing steps also offer better interpretability and ease for interactive error correction. Extensive human evaluations further confirm its effectiveness.

arXiv:2301.00152공식 코드 · Papers with Code

Towards Proactively Forecasting Sentence-Specific Information Popularity within Online News Documents

Multiple studies have focused on predicting the prospective popularity of an online document as a whole, without paying attention to the contributions of its individual parts. We introduce the task of proactively forecasting popularities of sentences within online news documents solely utilizing their natural language content. We model sentence-specific popularity forecasting as a sequence regression task. For training our models, we curate InfoPop, the first dataset containing popularity labels for over 1.7 million sentences from over 50,000 online news documents. To the best of our knowledge, this is the first dataset automatically created using streams of incoming search engine queries to generate sentence-level popularity annotations. We propose a novel transfer learning approach involving sentence salience prediction as an auxiliary task. Our proposed technique coupled with a BERT-based neural model exceeds nDCG values of 0.8 for proactive sentence-specific popularity forecasting. Notably, our study presents a non-trivial takeaway: though popularity and salience are different concepts, transfer learning from salience prediction enhances popularity forecasting. We release InfoPop and make our code publicly available: https://github.com/sayarghoshroy/InfoPopularity

arXiv:2212.14227공식 코드 · Papers with Code

StyleTTS-VC: One-Shot Voice Conversion by Knowledge Transfer from Style-Based TTS Models

One-shot voice conversion (VC) aims to convert speech from any source speaker to an arbitrary target speaker with only a few seconds of reference speech from the target speaker. This relies heavily on disentangling the speaker's identity and speech content, a task that still remains challenging. Here, we propose a novel approach to learning disentangled speech representation by transfer learning from style-based text-to-speech (TTS) models. With cycle consistent and adversarial training, the style-based TTS models can perform transcription-guided one-shot VC with high fidelity and similarity. By learning an additional mel-spectrogram encoder through a teacher-student knowledge transfer and novel data augmentation scheme, our approach results in disentangled speech representation without needing the input text. The subjective evaluation shows that our approach can significantly outperform the previous state-of-the-art one-shot voice conversion models in both naturalness and similarity.

arXiv:2212.14652

Domain-specific transfer learning in the automated scoring of tumor-stroma ratio from histopathological images of colorectal cancer

Tumor-stroma ratio (TSR) is a prognostic factor for many types of solid tumors. In this study, we propose a method for automated estimation of TSR from histopathological images of colorectal cancer. The method is based on convolutional neural networks which were trained to classify colorectal cancer tissue in hematoxylin-eosin stained samples into three classes: stroma, tumor and other. The models were trained using a data set that consists of 1343 whole slide images. Three different training setups were applied with a transfer learning approach using domain-specific data i.e. an external colorectal cancer histopathological data set. The three most accurate models were chosen as a classifier, TSR values were predicted and the results were compared to a visual TSR estimation made by a pathologist. The results suggest that classification accuracy does not improve when domain-specific data are used in the pre-training of the convolutional neural network models in the task at hand. Classification accuracy for stroma, tumor and other reached 96.1$\%$ on an independent test set. Among the three classes the best model gained the highest accuracy (99.3$\%$) for class tumor. When TSR was predicted with the best model, the correlation between the predicted values and values estimated by an experienced pathologist was 0.57. Further research is needed to study associations between computationally predicted TSR values and other clinicopathological factors of colorectal cancer and the overall survival of the patients.

arXiv:2212.14615

DRG-Net: Interactive Joint Learning of Multi-lesion Segmentation and Classification for Diabetic Retinopathy Grading

Diabetic Retinopathy (DR) is a leading cause of vision loss in the world, and early DR detection is necessary to prevent vision loss and support an appropriate treatment. In this work, we leverage interactive machine learning and introduce a joint learning framework, termed DRG-Net, to effectively learn both disease grading and multi-lesion segmentation. Our DRG-Net consists of two modules: (i) DRG-AI-System to classify DR Grading, localize lesion areas, and provide visual explanations; (ii) DRG-Expert-Interaction to receive feedback from user-expert and improve the DRG-AI-System. To deal with sparse data, we utilize transfer learning mechanisms to extract invariant feature representations by using Wasserstein distance and adversarial learning-based entropy minimization. Besides, we propose a novel attention strategy at both low- and high-level features to automatically select the most significant lesion information and provide explainable properties. In terms of human interaction, we further develop DRG-Net as a tool that enables expert users to correct the system's predictions, which may then be used to update the system as a whole. Moreover, thanks to the attention mechanism and loss functions constraint between lesion features and classification features, our approach can be robust given a certain level of noise in the feedback of users. We have benchmarked DRG-Net on the two largest DR datasets, i.e., IDRID and FGADR, and compared it to various state-of-the-art deep learning networks. In addition to outperforming other SOTA approaches, DRG-Net is effectively updated using user feedback, even in a weakly-supervised manner.

arXiv:2212.14278

Industrial Scene Change Detection using Deep Convolutional Neural Networks

Finding and localizing the conceptual changes in two scenes in terms of the presence or removal of objects in two images belonging to the same scene at different times in special care applications is of great significance. This is mainly due to the fact that addition or removal of important objects for some environments can be harmful. As a result, there is a need to design a program that locates these differences using machine vision. The most important challenge of this problem is the change in lighting conditions and the presence of shadows in the scene. Therefore, the proposed methods must be resistant to these challenges. In this article, a method based on deep convolutional neural networks using transfer learning is introduced, which is trained with an intelligent data synthesis process. The results of this method are tested and presented on the dataset provided for this purpose. It is shown that the presented method is more efficient than other methods and can be used in a variety of real industrial environments.

arXiv:2212.14040

HeartBEiT: Vision Transformer for Electrocardiogram Data Improves Diagnostic Performance at Low Sample Sizes

The electrocardiogram (ECG) is a ubiquitous diagnostic modality. Convolutional neural networks (CNNs) applied towards ECG analysis require large sample sizes, and transfer learning approaches result in suboptimal performance when pre-training is done on natural images. We leveraged masked image modeling to create the first vision-based transformer model, HeartBEiT, for electrocardiogram waveform analysis. We pre-trained this model on 8.5 million ECGs and then compared performance vs. standard CNN architectures for diagnosis of hypertrophic cardiomyopathy, low left ventricular ejection fraction and ST elevation myocardial infarction using differing training sample sizes and independent validation datasets. We show that HeartBEiT has significantly higher performance at lower sample sizes compared to other models. Finally, we also show that HeartBEiT improves explainability of diagnosis by highlighting biologically relevant regions of the EKG vs. standard CNNs. Thus, we present the first vision-based waveform transformer that can be used to develop specialized models for ECG analysis especially at low sample sizes.

arXiv:2212.13971

CT-LungNet: A Deep Learning Framework for Precise Lung Tissue Segmentation in 3D Thoracic CT Scans

Segmentation of lung tissue in computed tomography (CT) images is a precursor to most pulmonary image analysis applications. Semantic segmentation methods using deep learning have exhibited top-tier performance in recent years, however designing accurate and robust segmentation models for lung tissue is challenging due to the variations in shape, size, and orientation. Additionally, medical image artifacts and noise can affect lung tissue segmentation and degrade the accuracy of downstream analysis. The practicality of current deep learning methods for lung tissue segmentation is limited as they require significant computational resources and may not be easily deployable in clinical settings. This paper presents a fully automatic method that identifies the lungs in three-dimensional (3D) pulmonary CT images using deep networks and transfer learning. We introduce (1) a novel 2.5-dimensional image representation from consecutive CT slices that succinctly represents volumetric information and (2) a U-Net architecture equipped with pre-trained InceptionV3 blocks to segment 3D CT scans while maintaining the number of learnable parameters as low as possible. Our method was quantitatively assessed using one public dataset, LUNA16, for training and testing and two public datasets, namely, VESSEL12 and CRPF, only for testing. Due to the low number of learnable parameters, our method achieved high generalizability to the unseen VESSEL12 and CRPF datasets while obtaining superior performance over Luna16 compared to existing methods (Dice coefficients of 99.7, 99.1, and 98.8 over LUNA16, VESSEL12, and CRPF datasets, respectively). We made our method publicly accessible via a graphical user interface at medvispy.ee.kntu.ac.ir.

arXiv:2212.13970

Breaking the Architecture Barrier: A Method for Efficient Knowledge Transfer Across Networks

Transfer learning is a popular technique for improving the performance of neural networks. However, existing methods are limited to transferring parameters between networks with same architectures. We present a method for transferring parameters between neural networks with different architectures. Our method, called DPIAT, uses dynamic programming to match blocks and layers between architectures and transfer parameters efficiently. Compared to existing parameter prediction and random initialization methods, it significantly improves training efficiency and validation accuracy. In experiments on ImageNet, our method improved validation accuracy by an average of 1.6 times after 50 epochs of training. DPIAT allows both researchers and neural architecture search systems to modify trained networks and reuse knowledge, avoiding the need for retraining from scratch. We also introduce a network architecture similarity measure, enabling users to choose the best source network without any training.

arXiv:2212.13805공식 코드 · Papers with Code

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.

arXiv:2212.13591

Knowledge-Guided Data-Centric AI in Healthcare: Progress, Shortcomings, and Future Directions

The success of deep learning is largely due to the availability of large amounts of training data that cover a wide range of examples of a particular concept or meaning. In the field of medicine, having a diverse set of training data on a particular disease can lead to the development of a model that is able to accurately predict the disease. However, despite the potential benefits, there have not been significant advances in image-based diagnosis due to a lack of high-quality annotated data. This article highlights the importance of using a data-centric approach to improve the quality of data representations, particularly in cases where the available data is limited. To address this "small-data" issue, we discuss four methods for generating and aggregating training data: data augmentation, transfer learning, federated learning, and GANs (generative adversarial networks). We also propose the use of knowledge-guided GANs to incorporate domain knowledge in the training data generation process. With the recent progress in large pre-trained language models, we believe it is possible to acquire high-quality knowledge that can be used to improve the effectiveness of knowledge-guided generative methods.

arXiv:2212.12965

BD-KD: Balancing the Divergences for Online Knowledge Distillation

We address the challenge of producing trustworthy and accurate compact models for edge devices. While Knowledge Distillation (KD) has improved model compression in terms of achieving high accuracy performance, calibration of these compact models has been overlooked. We introduce BD-KD (Balanced Divergence Knowledge Distillation), a framework for logit-based online KD. BD-KD enhances both accuracy and model calibration simultaneously, eliminating the need for post-hoc recalibration techniques, which add computational overhead to the overall training pipeline and degrade performance. Our method encourages student-centered training by adjusting the conventional online distillation loss on both the student and teacher losses, employing sample-wise weighting of forward and reverse Kullback-Leibler divergence. This strategy balances student network confidence and boosts performance. Experiments across CIFAR10, CIFAR100, TinyImageNet, and ImageNet datasets, and various architectures demonstrate improved calibration and accuracy compared to recent online KD methods.