PyTorchKR Papers

최신 논문

총 10,240편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2403.06813

LeOCLR: Leveraging Original Images for Contrastive Learning of Visual Representations

Contrastive instance discrimination methods outperform supervised learning in downstream tasks such as image classification and object detection. However, these methods rely heavily on data augmentation during representation learning, which can lead to suboptimal results if not implemented carefully. A common augmentation technique in contrastive learning is random cropping followed by resizing. This can degrade the quality of representation learning when the two random crops contain distinct semantic content. To tackle this issue, we introduce LeOCLR (Leveraging Original Images for Contrastive Learning of Visual Representations), a framework that employs a novel instance discrimination approach and an adapted loss function. This method prevents the loss of important semantic features caused by mapping different object parts during representation learning. Our experiments demonstrate that LeOCLR consistently improves representation learning across various datasets, outperforming baseline models. For instance, LeOCLR surpasses MoCo-v2 by 5.1% on ImageNet-1K in linear evaluation and outperforms several other methods on transfer learning and object detection tasks.

arXiv:2403.06407공식 코드 · Papers with Code

Can LLMs' Tuning Methods Work in Medical Multimodal Domain?

While Large Language Models (LLMs) excel in world knowledge understanding, adapting them to specific subfields requires precise adjustments. Due to the model's vast scale, traditional global fine-tuning methods for large models can be computationally expensive and impact generalization. To address this challenge, a range of innovative Parameters-Efficient Fine-Tuning (PEFT) methods have emerged and achieved remarkable success in both LLMs and Large Vision-Language Models (LVLMs). In the medical domain, fine-tuning a medical Vision-Language Pretrained (VLP) model is essential for adapting it to specific tasks. Can the fine-tuning methods for large models be transferred to the medical field to enhance transfer learning efficiency? In this paper, we delve into the fine-tuning methods of LLMs and conduct extensive experiments to investigate the impact of fine-tuning methods for large models on the existing multimodal model in the medical domain from the training data level and the model structure level. We show the different impacts of fine-tuning methods for large models on medical VLMs and develop the most efficient ways to fine-tune medical VLP models. We hope this research can guide medical domain researchers in optimizing VLMs' training costs, fostering the broader application of VLMs in healthcare fields. The code and dataset have been released at https://github.com/TIMMY-CHAN/MILE.

arXiv:2403.06396

A Segmentation Foundation Model for Diverse-type Tumors

Large pre-trained models with their numerous model parameters and extensive training datasets have shown excellent performance in various tasks. Many publicly available medical image datasets do not have a sufficient amount of data so there are few large-scale models in medical imaging. We propose a large-scale Tumor Segmentation Foundation Model (TSFM) with 1.6 billion parameters using Resblock-backbone and Transformer-bottleneck,which has good transfer ability for downstream tasks. To make TSFM exhibit good performance in tumor segmentation, we make full use of the strong spatial correlation between tumors and organs in the medical image, innovatively fuse 7 tumor datasets and 3 multi-organ datasets to build a 3D medical dataset pool, including 2779 cases with totally 300k medical images, whose size currently exceeds many other single publicly available datasets. TSFM is the pre-trained model for medical image segmentation, which also can be transferred to multiple downstream tasks for fine-tuning learning. The average performance of our pre-trained model is 2% higher than that of nnU-Net across various tumor types. In the transfer learning task, TSFM only needs 5% training epochs of nnU-Net to achieve similar performance and can surpass nnU-Net by 2% on average with 10% training epoch. Pre-trained TSFM and its code will be released soon.

arXiv:2403.06108

Large Language Models on Fine-grained Emotion Detection Dataset with Data Augmentation and Transfer Learning

This paper delves into enhancing the classification performance on the GoEmotions dataset, a large, manually annotated dataset for emotion detection in text. The primary goal of this paper is to address the challenges of detecting subtle emotions in text, a complex issue in Natural Language Processing (NLP) with significant practical applications. The findings offer valuable insights into addressing the challenges of emotion detection in text and suggest directions for future research, including the potential for a survey paper that synthesizes methods and performances across various datasets in this domain.

arXiv:2403.06088

Towards In-Vehicle Multi-Task Facial Attribute Recognition: Investigating Synthetic Data and Vision Foundation Models

In the burgeoning field of intelligent transportation systems, enhancing vehicle-driver interaction through facial attribute recognition, such as facial expression, eye gaze, age, etc., is of paramount importance for safety, personalization, and overall user experience. However, the scarcity of comprehensive large-scale, real-world datasets poses a significant challenge for training robust multi-task models. Existing literature often overlooks the potential of synthetic datasets and the comparative efficacy of state-of-the-art vision foundation models in such constrained settings. This paper addresses these gaps by investigating the utility of synthetic datasets for training complex multi-task models that recognize facial attributes of passengers of a vehicle, such as gaze plane, age, and facial expression. Utilizing transfer learning techniques with both pre-trained Vision Transformer (ViT) and Residual Network (ResNet) models, we explore various training and adaptation methods to optimize performance, particularly when data availability is limited. We provide extensive post-evaluation analysis, investigating the effects of synthetic data distributions on model performance in in-distribution data and out-of-distribution inference. Our study unveils counter-intuitive findings, notably the superior performance of ResNet over ViTs in our specific multi-task context, which is attributed to the mismatch in model complexity relative to task complexity. Our results highlight the challenges and opportunities for enhancing the use of synthetic data and vision foundation models in practical applications.

arXiv:2403.06872공식 코드 · Papers with Code

Exploring Large Language Models and Hierarchical Frameworks for Classification of Large Unstructured Legal Documents

Legal judgment prediction suffers from the problem of long case documents exceeding tens of thousands of words, in general, and having a non-uniform structure. Predicting judgments from such documents becomes a challenging task, more so on documents with no structural annotation. We explore the classification of these large legal documents and their lack of structural information with a deep-learning-based hierarchical framework which we call MESc; "Multi-stage Encoder-based Supervised with-clustering"; for judgment prediction. Specifically, we divide a document into parts to extract their embeddings from the last four layers of a custom fine-tuned Large Language Model, and try to approximate their structure through unsupervised clustering. Which we use in another set of transformer encoder layers to learn the inter-chunk representations. We analyze the adaptability of Large Language Models (LLMs) with multi-billion parameters (GPT-Neo, and GPT-J) with the hierarchical framework of MESc and compare them with their standalone performance on legal texts. We also study their intra-domain(legal) transfer learning capability and the impact of combining embeddings from their last layers in MESc. We test these methods and their effectiveness with extensive experiments and ablation studies on legal documents from India, the European Union, and the United States with the ILDC dataset and a subset of the LexGLUE dataset. Our approach achieves a minimum total performance gain of approximately 2 points over previous state-of-the-art methods.

arXiv:2403.06621

Forest Inspection Dataset for Aerial Semantic Segmentation and Depth Estimation

Humans use UAVs to monitor changes in forest environments since they are lightweight and provide a large variety of surveillance data. However, their information does not present enough details for understanding the scene which is needed to assess the degree of deforestation. Deep learning algorithms must be trained on large amounts of data to output accurate interpretations, but ground truth recordings of annotated forest imagery are not available. To solve this problem, we introduce a new large aerial dataset for forest inspection which contains both real-world and virtual recordings of natural environments, with densely annotated semantic segmentation labels and depth maps, taken in different illumination conditions, at various altitudes and recording angles. We test the performance of two multi-scale neural networks for solving the semantic segmentation task (HRNet and PointFlow network), studying the impact of the various acquisition conditions and the capabilities of transfer learning from virtual to real data. Our results showcase that the best results are obtained when the training is done on a dataset containing a large variety of scenarios, rather than separating the data into specific categories. We also develop a framework to assess the deforestation degree of an area.

arXiv:2403.06601공식 코드 · Papers with Code

Cross-domain and Cross-dimension Learning for Image-to-Graph Transformers

Direct image-to-graph transformation is a challenging task that involves solving object detection and relationship prediction in a single model. Due to this task's complexity, large training datasets are rare in many domains, making the training of deep-learning methods challenging. This data sparsity necessitates transfer learning strategies akin to the state-of-the-art in general computer vision. In this work, we introduce a set of methods enabling cross-domain and cross-dimension learning for image-to-graph transformers. We propose (1) a regularized edge sampling loss to effectively learn object relations in multiple domains with different numbers of edges, (2) a domain adaptation framework for image-to-graph transformers aligning image- and graph-level features from different domains, and (3) a projection function that allows using 2D data for training 3D transformers. We demonstrate our method's utility in cross-domain and cross-dimension experiments, where we utilize labeled data from 2D road networks for simultaneous learning in vastly different target domains. Our method consistently outperforms standard transfer learning and self-supervised pretraining on challenging benchmarks, such as retinal or whole-brain vessel graph extraction.

arXiv:2403.06382

Pre-Trained Model Recommendation for Downstream Fine-tuning

As a fundamental problem in transfer learning, model selection aims to rank off-the-shelf pre-trained models and select the most suitable one for the new target task. Existing model selection techniques are often constrained in their scope and tend to overlook the nuanced relationships between models and tasks. In this paper, we present a pragmatic framework \textbf{Fennec}, delving into a diverse, large-scale model repository while meticulously considering the intricate connections between tasks and models. The key insight is to map all models and historical tasks into a transfer-related subspace, where the distance between model vectors and task vectors represents the magnitude of transferability. A large vision model, as a proxy, infers a new task's representation in the transfer space, thereby circumventing the computational burden of extensive forward passes. We also investigate the impact of the inherent inductive bias of models on transfer results and propose a novel method called \textbf{archi2vec} to encode the intricate structures of models. The transfer score is computed through straightforward vector arithmetic with a time complexity of $\mathcal{O}(1)$. Finally, we make a substantial contribution to the field by releasing a comprehensive benchmark. We validate the effectiveness of our framework through rigorous testing on two benchmarks. The benchmark and the code will be publicly available in the near future.

arXiv:2403.05519

Authorship Attribution in Bangla Literature (AABL) via Transfer Learning using ULMFiT

Authorship Attribution is the task of creating an appropriate characterization of text that captures the authors' writing style to identify the original author of a given piece of text. With increased anonymity on the internet, this task has become increasingly crucial in various security and plagiarism detection fields. Despite significant advancements in other languages such as English, Spanish, and Chinese, Bangla lacks comprehensive research in this field due to its complex linguistic feature and sentence structure. Moreover, existing systems are not scalable when the number of author increases, and the performance drops for small number of samples per author. In this paper, we propose the use of Average-Stochastic Gradient Descent Weight-Dropped Long Short-Term Memory (AWD-LSTM) architecture and an effective transfer learning approach that addresses the problem of complex linguistic features extraction and scalability for authorship attribution in Bangla Literature (AABL). We analyze the effect of different tokenization, such as word, sub-word, and character level tokenization, and demonstrate the effectiveness of these tokenizations in the proposed model. Moreover, we introduce the publicly available Bangla Authorship Attribution Dataset of 16 authors (BAAD16) containing 17,966 sample texts and 13.4+ million words to solve the standard dataset scarcity problem and release six variations of pre-trained language models for use in any Bangla NLP downstream task. For evaluation, we used our developed BAAD16 dataset as well as other publicly available datasets. Empirically, our proposed model outperformed state-of-the-art models and achieved 99.8% accuracy in the BAAD16 dataset. Furthermore, we showed that the proposed system scales much better even with an increasing number of authors, and performance remains steady despite few training samples.

arXiv:2403.04837공식 코드 · Papers with Code

Cell reprogramming design by transfer learning of functional transcriptional networks

Recent developments in synthetic biology, next-generation sequencing, and machine learning provide an unprecedented opportunity to rationally design new disease treatments based on measured responses to gene perturbations and drugs to reprogram cells. The main challenges to seizing this opportunity are the incomplete knowledge of the cellular network and the combinatorial explosion of possible interventions, both of which are insurmountable by experiments. To address these challenges, we develop a transfer learning approach to control cell behavior that is pre-trained on transcriptomic data associated with human cell fates, thereby generating a model of the network dynamics that can be transferred to specific reprogramming goals. The approach combines transcriptional responses to gene perturbations to minimize the difference between a given pair of initial and target transcriptional states. We demonstrate our approach's versatility by applying it to a microarray dataset comprising >9,000 microarrays across 54 cell types and 227 unique perturbations, and an RNASeq dataset consisting of >10,000 sequencing runs across 36 cell types and 138 perturbations. Our approach reproduces known reprogramming protocols with an AUROC of 0.91 while innovating over existing methods by pre-training an adaptable model that can be tailored to specific reprogramming transitions. We show that the number of gene perturbations required to steer from one fate to another increases with decreasing developmental relatedness and that fewer genes are needed to progress along developmental paths than to regress. These findings establish a proof-of-concept for our approach to computationally design control strategies and provide insights into how gene regulatory networks govern phenotype.

arXiv:2403.05489공식 코드 · Papers with Code

JointMotion: Joint Self-Supervision for Joint Motion Prediction

We present JointMotion, a self-supervised pre-training method for joint motion prediction in self-driving vehicles. Our method jointly optimizes a scene-level objective connecting motion and environments, and an instance-level objective to refine learned representations. Scene-level representations are learned via non-contrastive similarity learning of past motion sequences and environment context. At the instance level, we use masked autoencoding to refine multimodal polyline representations. We complement this with an adaptive pre-training decoder that enables JointMotion to generalize across different environment representations, fusion mechanisms, and dataset characteristics. Notably, our method reduces the joint final displacement error of Wayformer, HPTR, and Scene Transformer models by 3\%, 8\%, and 12\%, respectively; and enables transfer learning between the Waymo Open Motion and the Argoverse 2 Motion Forecasting datasets. Code: https://github.com/kit-mrt/future-motion

arXiv:2403.05353

Hybridized Convolutional Neural Networks and Long Short-Term Memory for Improved Alzheimer's Disease Diagnosis from MRI Scans

Brain-related diseases are more sensitive than other diseases due to several factors, including the complexity of surgical procedures, high costs, and other challenges. Alzheimer's disease is a common brain disorder that causes memory loss and the shrinking of brain cells. Early detection is critical for providing proper treatment to patients. However, identifying Alzheimer's at an early stage using manual scanning of CT or MRI scans is challenging. Therefore, researchers have delved into the exploration of computer-aided systems, employing Machine Learning and Deep Learning methodologies, which entail the training of datasets to detect Alzheimer's disease. This study aims to present a hybrid model that combines a CNN model's feature extraction capabilities with an LSTM model's detection capabilities. This study has applied the transfer learning called VGG16 in the hybrid model to extract features from MRI images. The LSTM detects features between the convolution layer and the fully connected layer. The output layer of the fully connected layer uses the softmax function. The training of the hybrid model involved utilizing the ADNI dataset. The trial findings revealed that the model achieved a level of accuracy of 98.8%, a sensitivity rate of 100%, and a specificity rate of 76%. The proposed hybrid model outperforms its contemporary CNN counterparts, showcasing a superior performance.

arXiv:2403.05061공식 코드 · Papers with Code

RadarDistill: Boosting Radar-based Object Detection Performance via Knowledge Distillation from LiDAR Features

The inherent noisy and sparse characteristics of radar data pose challenges in finding effective representations for 3D object detection. In this paper, we propose RadarDistill, a novel knowledge distillation (KD) method, which can improve the representation of radar data by leveraging LiDAR data. RadarDistill successfully transfers desirable characteristics of LiDAR features into radar features using three key components: Cross-Modality Alignment (CMA), Activation-based Feature Distillation (AFD), and Proposal-based Feature Distillation (PFD). CMA enhances the density of radar features by employing multiple layers of dilation operations, effectively addressing the challenge of inefficient knowledge transfer from LiDAR to radar. AFD selectively transfers knowledge based on regions of the LiDAR features, with a specific focus on areas where activation intensity exceeds a predefined threshold. PFD similarly guides the radar network to selectively mimic features from the LiDAR network within the object proposals. Our comparative analyses conducted on the nuScenes datasets demonstrate that RadarDistill achieves state-of-the-art (SOTA) performance for radar-only object detection task, recording 20.5% in mAP and 43.7% in NDS. Also, RadarDistill significantly improves the performance of the camera-radar fusion model.

arXiv:2403.05396공식 코드 · Papers with Code

HistGen: Histopathology Report Generation via Local-Global Feature Encoding and Cross-modal Context Interaction

Histopathology serves as the gold standard in cancer diagnosis, with clinical reports being vital in interpreting and understanding this process, guiding cancer treatment and patient care. The automation of histopathology report generation with deep learning stands to significantly enhance clinical efficiency and lessen the labor-intensive, time-consuming burden on pathologists in report writing. In pursuit of this advancement, we introduce HistGen, a multiple instance learning-empowered framework for histopathology report generation together with the first benchmark dataset for evaluation. Inspired by diagnostic and report-writing workflows, HistGen features two delicately designed modules, aiming to boost report generation by aligning whole slide images (WSIs) and diagnostic reports from local and global granularity. To achieve this, a local-global hierarchical encoder is developed for efficient visual feature aggregation from a region-to-slide perspective. Meanwhile, a cross-modal context module is proposed to explicitly facilitate alignment and interaction between distinct modalities, effectively bridging the gap between the extensive visual sequences of WSIs and corresponding highly summarized reports. Experimental results on WSI report generation show the proposed model outperforms state-of-the-art (SOTA) models by a large margin. Moreover, the results of fine-tuning our model on cancer subtyping and survival analysis tasks further demonstrate superior performance compared to SOTA methods, showcasing strong transfer learning capability. Dataset, model weights, and source code are available in https://github.com/dddavid4real/HistGen.

arXiv:2403.05260공식 코드 · Papers with Code

Towards generalization of drug response prediction to single cells and patients utilizing importance-aware multi-source domain transfer learning

The advancement of single-cell sequencing technology has promoted the generation of a large amount of single-cell transcriptional profiles, providing unprecedented opportunities to identify drug-resistant cell subpopulations within a tumor. However, few studies have focused on drug response prediction at single-cell level, and their performance remains suboptimal. This paper proposed scAdaDrug, a novel multi-source domain adaptation model powered by adaptive importance-aware representation learning to predict drug response of individual cells. We used a shared encoder to extract domain-invariant features related to drug response from multiple source domains by utilizing adversarial domain adaptation. Particularly, we introduced a plug-and-play module to generate importance-aware and mutually independent weights, which could adaptively modulate the latent representation of each sample in element-wise manner between source and target domains. Extensive experimental results showed that our model achieved state-of-the-art performance in predicting drug response on multiple independent datasets, including single-cell datasets derived from both cell lines and patient-derived xenografts (PDX) models, as well as clinical tumor patient cohorts. Moreover, the ablation experiments demonstrated our model effectively captured the underlying patterns determining drug response from multiple source domains.

arXiv:2403.05164공식 코드 · Papers with Code

Synthetic data generation for system identification: leveraging knowledge transfer from similar systems

This paper addresses the challenge of overfitting in the learning of dynamical systems by introducing a novel approach for the generation of synthetic data, aimed at enhancing model generalization and robustness in scenarios characterized by data scarcity. Central to the proposed methodology is the concept of knowledge transfer from systems within the same class. Specifically, synthetic data is generated through a pre-trained meta-model that describes a broad class of systems to which the system of interest is assumed to belong. Training data serves a dual purpose: firstly, as input to the pre-trained meta model to discern the system's dynamics, enabling the prediction of its behavior and thereby generating synthetic output sequences for new input sequences; secondly, in conjunction with synthetic data, to define the loss function used for model estimation. A validation dataset is used to tune a scalar hyper-parameter balancing the relative importance of training and synthetic data in the definition of the loss function. The same validation set can be also used for other purposes, such as early stopping during the training, fundamental to avoid overfitting in case of small-size training datasets. The efficacy of the approach is shown through a numerical example that highlights the advantages of integrating synthetic data into the system identification process.

논문 ID: document/9548667

QBox: Partial Transfer Learning with Active Querying for Object Detection

Object detection requires plentiful data annotated with bounding boxes for model training. However, in many applications, it is difficult or even impossible to acquire a large set of labeled examples for the target task due to the privacy concern or lack of reliable annotators. On the other hand, due to the high-quality image search engines, such as Flickr and Google , it is relatively easy to obtain resource-rich unlabeled datasets, whose categories are a superset of those of target data. In this article, to improve the target model with cost-effective supervision from source data, we propose a partial transfer learning approach QBox to actively query labels for bounding boxes of source images. Specifically, we design two criteria, i.e., informativeness and transferability, to measure the potential utility of a bounding box for improving the target model. Based on these criteria, QBox actively queries the labels of the most useful boxes from the source domain and, thus, requires fewer training examples to save the labeling cost. Furthermore, the proposed query strategy allows annotators to simply labeling a specific region, instead of the whole image, and, thus, significantly reduces the labeling difficulty. Extensive experiments are performed on various partial transfer benchmarks and a real COVID-19 detection task. The results validate that QBox improves the detection accuracy with lower labeling cost compared to state-of-the-art query strategies for object detection.

arXiv:2403.04484공식 코드 · Papers with Code

Source Matters: Source Dataset Impact on Model Robustness in Medical Imaging

Transfer learning has become an essential part of medical imaging classification algorithms, often leveraging ImageNet weights. The domain shift from natural to medical images has prompted alternatives such as RadImageNet, often showing comparable classification performance. However, it remains unclear whether the performance gains from transfer learning stem from improved generalization or shortcut learning. To address this, we conceptualize confounders by introducing the Medical Imaging Contextualized Confounder Taxonomy (MICCAT) and investigate a range of confounders across it -- whether synthetic or sampled from the data -- using two public chest X-ray and CT datasets. We show that ImageNet and RadImageNet achieve comparable classification performance, yet ImageNet is much more prone to overfitting to confounders. We recommend that researchers using ImageNet-pretrained models reexamine their model robustness by conducting similar experiments. Our code and experiments are available at https://github.com/DovileDo/source-matters.

arXiv:2403.04697공식 코드 · Papers with Code

AUFormer: Vision Transformers are Parameter-Efficient Facial Action Unit Detectors

Facial Action Units (AU) is a vital concept in the realm of affective computing, and AU detection has always been a hot research topic. Existing methods suffer from overfitting issues due to the utilization of a large number of learnable parameters on scarce AU-annotated datasets or heavy reliance on substantial additional relevant data. Parameter-Efficient Transfer Learning (PETL) provides a promising paradigm to address these challenges, whereas its existing methods lack design for AU characteristics. Therefore, we innovatively investigate PETL paradigm to AU detection, introducing AUFormer and proposing a novel Mixture-of-Knowledge Expert (MoKE) collaboration mechanism. An individual MoKE specific to a certain AU with minimal learnable parameters first integrates personalized multi-scale and correlation knowledge. Then the MoKE collaborates with other MoKEs in the expert group to obtain aggregated information and inject it into the frozen Vision Transformer (ViT) to achieve parameter-efficient AU detection. Additionally, we design a Margin-truncated Difficulty-aware Weighted Asymmetric Loss (MDWA-Loss), which can encourage the model to focus more on activated AUs, differentiate the difficulty of unactivated AUs, and discard potential mislabeled samples. Extensive experiments from various perspectives, including within-domain, cross-domain, data efficiency, and micro-expression domain, demonstrate AUFormer's state-of-the-art performance and robust generalization abilities without relying on additional relevant data. The code for AUFormer is available at https://github.com/yuankaishen2001/AUFormer.