Combining electroencephalogram (EEG) datasets for supervised machine learning (ML) is challenging due to session, subject, and device variability. ML algorithms typically require identical features at train and test time, complicating analysis due to varying sensor numbers and positions across datasets. Simple channel selection discards valuable data, leading to poorer performance, especially with datasets sharing few channels. To address this, we propose an unsupervised approach leveraging EEG signal physics. We map EEG channels to fixed positions using field interpolation, facilitating source-free domain adaptation. Leveraging Riemannian geometry classification pipelines and transfer learning steps, our method demonstrates robust performance in brain-computer interface (BCI) tasks and potential biomarker applications. Comparative analysis against a statistical-based approach known as Dimensionality Transcending, a signal-based imputation called ComImp, source-dependent methods, as well as common channel selection and spherical spline interpolation, was conducted with leave-one-dataset-out validation on six public BCI datasets for a right-hand/left-hand classification task. Numerical experiments show that in the presence of few shared channels in train and test, the field interpolation consistently outperforms other methods, demonstrating enhanced classification performance across all datasets. When more channels are shared, field interpolation was found to be competitive with other methods and faster to compute than source-dependent methods.
This paper proposes a novel localization algorithm using the reconfigurable intelligent surface (RIS) received signal, i.e., RIS information. Compared with BS received signal, i.e., BS information, RIS information offers higher dimension and richer feature set, thereby providing an enhanced capacity to distinguish positions of the mobile users (MUs). Additionally, we address a practical scenario where RIS contains some unknown (number and places) faulty elements that cannot receive signals. Initially, we employ transfer learning to design a two-phase transfer learning (TPTL) algorithm, designed for accurate detection of faulty elements. Then our objective is to regain the information lost from the faulty elements and reconstruct the complete high-dimensional RIS information for localization. To this end, we propose a transfer-enhanced dual-stage (TEDS) algorithm. In \emph{Stage I}, we integrate the CNN and variational autoencoder (VAE) to obtain the RIS information, which in \emph{Stage II}, is input to the transferred DenseNet 121 to estimate the location of the MU. To gain more insight, we propose an alternative algorithm named transfer-enhanced direct fingerprint (TEDF) algorithm which only requires the BS information. The comparison between TEDS and TEDF reveals the effectiveness of faulty element detection and the benefits of utilizing the high-dimensional RIS information for localization. Besides, our empirical results demonstrate that the performance of the localization algorithm is dominated by the high-dimensional RIS information and is robust to unoptimized phase shifts and signal-to-noise ratio (SNR).
Reconfigurable intelligent surface (RIS)-aided localization systems have attracted extensive research attention due to their accuracy enhancement capabilities. However, most studies primarily utilized the base stations (BS) received signal, i.e., BS information, for localization algorithm design, neglecting the potential of RIS received signal, i.e., RIS information. Compared with BS information, RIS information offers higher dimension and richer feature set, thereby significantly improving the ability to extract positions of the mobile users (MUs). Addressing this oversight, this paper explores the algorithm design based on the high-dimensional RIS information. Specifically, we first propose a RIS information reconstruction (RIS-IR) algorithm to reconstruct the high-dimensional RIS information from the low-dimensional BS information. The proposed RIS-IR algorithm comprises a data processing module for preprocessing BS information, a convolution neural network (CNN) module for feature extraction, and an output module for outputting the reconstructed RIS information. Then, we propose a transfer learning based fingerprint (TFBF) algorithm that employs the reconstructed high-dimensional RIS information for MU localization. This involves adapting a pre-trained DenseNet-121 model to map the reconstructed RIS signal to the MU's three-dimensional (3D) position. Empirical results affirm that the localization performance is significantly influenced by the high-dimensional RIS information and maintains robustness against unoptimized phase shifts.
Deep learning has become an extremely powerful tool for complex tasks such as image classification and segmentation. The medical industry often lacks high-quality, balanced datasets, which can be a challenge for deep learning algorithms that need sufficiently large amounts of data to train and increase their performance. This is especially important in the context of kidney issues such as for stones, cysts and tumors. We used deep learning models for this study to classify or detect several types of kidney diseases. We use different classification models, such as VGG-19, (CNNs) Convolutional Neural Networks, ResNet-101, VGG-16, ResNet-50, and DenseNet-169, which can be enhanced through techniques such as classification, segmentation, and transfer learning. These algorithms can help improve model accuracy by allowing them to learn from multiple datasets. This technique has the potential to revolutionize the diagnosis and treatment of kidney problems as it enables more accurate and effective classification of CT-scan images. This may ultimately lead to better patient outcomes and improved overall health outcomes.
Despite living in a multi-sensory world, most AI models are limited to textual and visual understanding of human motion and behavior. In fact, full situational awareness of human motion could best be understood through a combination of sensors. In this survey we investigate how knowledge can be transferred and utilized amongst modalities for Human Activity/Action Recognition (HAR), i.e. cross-modality transfer learning. We motivate the importance and potential of IMU data and its applicability in cross-modality learning as well as the importance of studying the HAR problem. We categorize HAR related tasks by time and abstractness and then compare various types of multimodal HAR datasets. We also distinguish and expound on many related but inconsistently used terms in the literature, such as transfer learning, domain adaptation, representation learning, sensor fusion, and multimodal learning, and describe how cross-modal learning fits with all these concepts. We then review the literature in IMU-based cross-modal transfer for HAR. The two main approaches for cross-modal transfer are instance-based transfer, where instances of one modality are mapped to another (e.g. knowledge is transferred in the input space), or feature-based transfer, where the model relates the modalities in an intermediate latent space (e.g. knowledge is transferred in the feature space). Finally, we discuss future research directions and applications in cross-modal HAR.
Text continues to remain a relevant form of representation for information. Text documents are created either in digital native platforms or through the conversion of other media files such as images and speech. While the digital native text is invariably obtained through physical or virtual keyboards, technologies such as OCR and speech recognition are utilized to transform the images and speech signals into text content. All these variety of mechanisms of text generation also introduce errors into the captured text. This project aims at analyzing different kinds of error that occurs in text documents. The work employs two of the advanced deep neural network-based language models, namely, BART and MarianMT, to rectify the anomalies present in the text. Transfer learning of these models with available dataset is performed to finetune their capacity for error correction. A comparative study is conducted to investigate the effectiveness of these models in handling each of the defined error categories. It is observed that while both models can bring down the erroneous sentences by 20+%, BART can handle spelling errors far better (24.6%) than grammatical errors (8.8%).
The success of Natural Language Understanding (NLU) benchmarks in various languages, such as GLUE for English, CLUE for Chinese, KLUE for Korean, and IndoNLU for Indonesian, has facilitated the evaluation of new NLU models across a wide range of tasks. To establish a standardized set of benchmarks for Vietnamese NLU, we introduce the first Vietnamese Language Understanding Evaluation (VLUE) benchmark. The VLUE benchmark encompasses five datasets covering different NLU tasks, including text classification, span extraction, and natural language understanding. To provide an insightful overview of the current state of Vietnamese NLU, we then evaluate seven state-of-the-art pre-trained models, including both multilingual and Vietnamese monolingual models, on our proposed VLUE benchmark. Furthermore, we present CafeBERT, a new state-of-the-art pre-trained model that achieves superior results across all tasks in the VLUE benchmark. Our model combines the proficiency of a multilingual pre-trained model with Vietnamese linguistic knowledge. CafeBERT is developed based on the XLM-RoBERTa model, with an additional pretraining step utilizing a significant amount of Vietnamese textual data to enhance its adaptation to the Vietnamese language. For the purpose of future research, CafeBERT is made publicly available for research purposes.
Addressing the challenge of data scarcity in industrial domains, transfer learning emerges as a pivotal paradigm. This work introduces Style Filter, a tailored methodology for industrial contexts. By selectively filtering source domain data before knowledge transfer, Style Filter reduces the quantity of data while maintaining or even enhancing the performance of transfer learning strategy. Offering label-free operation, minimal reliance on prior knowledge, independence from specific models, and re-utilization, Style Filter is evaluated on authentic industrial datasets, highlighting its effectiveness when employed before conventional transfer strategies in the deep learning domain. The results underscore the effectiveness of Style Filter in real-world industrial applications.
In the last few years, the artifact patterns in fake images synthesized by different generative models have been inconsistent, leading to the failure of previous research that relied on spotting subtle differences between real and fake. In our preliminary experiments, we find that the artifacts in fake images always change with the development of the generative model, while natural images exhibit stable statistical properties. In this paper, we employ natural traces shared only by real images as an additional target for a classifier. Specifically, we introduce a self-supervised feature mapping process for natural trace extraction and develop a transfer learning based on soft contrastive loss to bring them closer to real images and further away from fake ones. This motivates the detector to make decisions based on the proximity of images to the natural traces. To conduct a comprehensive experiment, we built a high-quality and diverse dataset that includes generative models comprising GANs and diffusion models, to evaluate the effectiveness in generalizing unknown forgery techniques and robustness in surviving different transformations. Experimental results show that our proposed method gives 96.2% mAP significantly outperforms the baselines. Extensive experiments conducted on popular commercial platforms reveal that our proposed method achieves an accuracy exceeding 78.4%, underscoring its practicality for real-world application deployment.
Self-supervised learning has emerged as a powerful tool for pretraining deep networks on unlabeled data, prior to transfer learning of target tasks with limited annotation. The relevance between the pretraining pretext and target tasks is crucial to the success of transfer learning. Various pretext tasks have been proposed to utilize properties of medical image data (e.g., three dimensionality), which are more relevant to medical image analysis than generic ones for natural images. However, previous work rarely paid attention to data with anatomy-oriented imaging planes, e.g., standard cardiac magnetic resonance imaging views. As these imaging planes are defined according to the anatomy of the imaged organ, pretext tasks effectively exploiting this information can pretrain the networks to gain knowledge on the organ of interest. In this work, we propose two complementary pretext tasks for this group of medical image data based on the spatial relationship of the imaging planes. The first is to learn the relative orientation between the imaging planes and implemented as regressing their intersecting lines. The second exploits parallel imaging planes to regress their relative slice locations within a stack. Both pretext tasks are conceptually straightforward and easy to implement, and can be combined in multitask learning for better representation learning. Thorough experiments on two anatomical structures (heart and knee) and representative target tasks (semantic segmentation and classification) demonstrate that the proposed pretext tasks are effective in pretraining deep networks for remarkably boosted performance on the target tasks, and superior to other recent approaches.
This paper tackles a novel yet challenging problem: how to transfer knowledge from the emerging Segment Anything Model (SAM) -- which reveals impressive zero-shot instance segmentation capacity -- to learn a compact panoramic semantic segmentation model, i.e., student, without requiring any labeled data. This poses considerable challenges due to SAM's inability to provide semantic labels and the large capacity gap between SAM and the student. To this end, we propose a novel framework, called GoodSAM, that introduces a teacher assistant (TA) to provide semantic information, integrated with SAM to generate ensemble logits to achieve knowledge transfer. Specifically, we propose a Distortion-Aware Rectification (DAR) module that first addresses the distortion problem of panoramic images by imposing prediction-level consistency and boundary enhancement. This subtly enhances TA's prediction capacity on panoramic images. DAR then incorporates a cross-task complementary fusion block to adaptively merge the predictions of SAM and TA to obtain more reliable ensemble logits. Moreover, we introduce a Multi-level Knowledge Adaptation (MKA) module to efficiently transfer the multi-level feature knowledge from TA and ensemble logits to learn a compact student model. Extensive experiments on two benchmarks show that our GoodSAM achieves a remarkable +3.75\% mIoU improvement over the state-of-the-art (SOTA) domain adaptation methods. Also, our most lightweight model achieves comparable performance to the SOTA methods with only 3.7M parameters.
Continual learning (CL) aims to empower models to learn new tasks without forgetting previously acquired knowledge. Most prior works concentrate on the techniques of architectures, replay data, regularization, \etc. However, the category name of each class is largely neglected. Existing methods commonly utilize the one-hot labels and randomly initialize the classifier head. We argue that the scarce semantic information conveyed by the one-hot labels hampers the effective knowledge transfer across tasks. In this paper, we revisit the role of the classifier head within the CL paradigm and replace the classifier with semantic knowledge from pretrained language models (PLMs). Specifically, we use PLMs to generate semantic targets for each class, which are frozen and serve as supervision signals during training. Such targets fully consider the semantic correlation between all classes across tasks. Empirical studies show that our approach mitigates forgetting by alleviating representation drifting and facilitating knowledge transfer across tasks. The proposed method is simple to implement and can seamlessly be plugged into existing methods with negligible adjustments. Extensive experiments based on eleven mainstream baselines demonstrate the effectiveness and generalizability of our approach to various protocols. For example, under the class-incremental learning setting on ImageNet-100, our method significantly improves the Top-1 accuracy by 3.2\% to 6.1\% while reducing the forgetting rate by 2.6\% to 13.1\%.
Recently, automated medical image segmentation methods based on deep learning have achieved great success. However, they heavily rely on large annotated datasets, which are costly and time-consuming to acquire. Few-shot learning aims to overcome the need for annotated data by using a small labeled dataset, known as a support set, to guide predicting labels for new, unlabeled images, known as the query set. Inspired by this paradigm, we introduce MatchSeg, a novel framework that enhances medical image segmentation through strategic reference image matching. We leverage contrastive language-image pre-training (CLIP) to select highly relevant samples when defining the support set. Additionally, we design a joint attention module to strengthen the interaction between support and query features, facilitating a more effective knowledge transfer between support and query sets. We validated our method across four public datasets. Experimental results demonstrate superior segmentation performance and powerful domain generalization ability of MatchSeg against existing methods for domain-specific and cross-domain segmentation tasks. Our code is made available at https://github.com/keeplearning-again/MatchSeg
Heterogeneous Federated Learning (HtFL) enables task-specific knowledge sharing among clients with different model architectures while preserving privacy. Despite recent research progress, transferring knowledge in HtFL is still difficult due to data and model heterogeneity. To tackle this, we introduce a public pre-trained generator (e.g., StyleGAN or Stable Diffusion) as the bridge and propose a new upload-efficient knowledge transfer scheme called Federated Knowledge-Transfer-Loop (FedKTL). It can produce task-related prototypical image-vector pairs via the generator's inference on the server. With these pairs, each client can transfer common knowledge from the generator to its local model through an additional supervised local task. We conduct extensive experiments on four datasets under two types of data heterogeneity with 14 heterogeneous models, including CNNs and ViTs. Results show that our FedKTL surpasses seven state-of-the-art methods by up to 7.31%. Moreover, our knowledge transfer scheme is applicable in cloud-edge scenarios with only one edge client. Code: https://github.com/TsingZ0/FedKTL
The Sentinel-2 satellite constellation produces high-resolution multispectral data, covering the entire Earth's land surface every five days. However, the use of this data is significantly impacted by the presence of pixels affected by cloud and shadow. Precise and efficient cloud and cloud shadow masking methods are required for the automated use of this data. Here, we present CloudS2Mask, a new and open-source Python deep-learning library to detect cloud and cloud shadows in Sentinel-2 data accurately and efficiently. CloudS2Mask is trained on multiple large global datasets including CloudSEN12, KappaSet, Sentinel-2 Cloud Mask Catalogue, Sentinel-2 reference cloud masks generated by an active learning method and a custom CloudS2Mask dataset. Using techniques such as progressive resizing, hard fine-tuning, and transfer learning, CloudS2Mask outperforms published models when evaluated on both the CloudSEN12 and the PixBox datasets in terms of classification accuracy. Compared to the next best-published results, CloudS2Mask improves Balanced Overall Accuracy metrics on the CloudSEN12 dataset by 1.2% for cloud, 2.8% for cloud shadow and 1.1% for clear classification. CloudS2Mask provides a range of customisable options combined with novel pre-processing and post-processing approaches to meet task-specific needs, enabling mean scene processing speeds up to 44 times faster than current state-of-the-art methods. The application of novel techniques in CloudS2Mask, combined with its accuracy and processing speed, underscores its potential to streamline the automated cloud and cloud shadow detection of Sentinel-2 data.
In this paper, we propose a novel parameter and computation efficient tuning method for Multi-modal Large Language Models (MLLMs), termed Efficient Attention Skipping (EAS). Concretely, we first reveal that multi-head attentions (MHAs), the main computational overhead of MLLMs, are often redundant to downstream tasks. Based on this observation, EAS evaluates the attention redundancy and skips the less important MHAs to speed up inference. Besides, we also propose a novel propagation-of-information adapter (PIA) to serve the attention skipping of EAS and keep parameter efficiency, which can be further re-parameterized into feed-forward networks (FFNs) for zero-extra latency. To validate EAS, we apply it to a recently proposed MLLM called LaVIN and a classic VL pre-trained model called METER, and conduct extensive experiments on a set of benchmarks. The experiments show that EAS not only retains high performance and parameter efficiency, but also greatly speeds up inference speed. For instance, LaVIN-EAS can obtain 89.98\% accuracy on ScineceQA while speeding up inference by 2.2 times to LaVIN
Transformers have widely adopted attention networks for sequence mixing and MLPs for channel mixing, playing a pivotal role in achieving breakthroughs across domains. However, recent literature highlights issues with attention networks, including low inductive bias and quadratic complexity concerning input sequence length. State Space Models (SSMs) like S4 and others (Hippo, Global Convolutions, liquid S4, LRU, Mega, and Mamba), have emerged to address the above issues to help handle longer sequence lengths. Mamba, while being the state-of-the-art SSM, has a stability issue when scaled to large networks for computer vision datasets. We propose SiMBA, a new architecture that introduces Einstein FFT (EinFFT) for channel modeling by specific eigenvalue computations and uses the Mamba block for sequence modeling. Extensive performance studies across image and time-series benchmarks demonstrate that SiMBA outperforms existing SSMs, bridging the performance gap with state-of-the-art transformers. Notably, SiMBA establishes itself as the new state-of-the-art SSM on ImageNet and transfer learning benchmarks such as Stanford Car and Flower as well as task learning benchmarks as well as seven time series benchmark datasets. The project page is available on this website ~\url{https://github.com/badripatro/Simba}.
This paper addresses the critical challenge of vehicle detection in the harsh winter conditions in the Nordic regions, characterized by heavy snowfall, reduced visibility, and low lighting. Due to their susceptibility to environmental distortions and occlusions, traditional vehicle detection methods have struggled in these adverse conditions. The advanced proposed deep learning architectures brought promise, yet the unique difficulties of detecting vehicles in Nordic winters remain inadequately addressed. This study uses the Nordic Vehicle Dataset (NVD), which has UAV images from northern Sweden, to evaluate the performance of state-of-the-art vehicle detection algorithms under challenging weather conditions. Our methodology includes a comprehensive evaluation of single-stage, two-stage, and transformer-based detectors against the NVD. We propose a series of enhancements tailored to each detection framework, including data augmentation, hyperparameter tuning, transfer learning, and novel strategies designed explicitly for the DETR model. Our findings not only highlight the limitations of current detection systems in the Nordic environment but also offer promising directions for enhancing these algorithms for improved robustness and accuracy in vehicle detection amidst the complexities of winter landscapes. The code and the dataset are available at https://nvd.ltu-ai.dev
As deep learning continues to evolve, we have observed huge breakthroughs in the fields of medical imaging, video and frame generation, optical character recognition (OCR), and other domains. In the field of data analysis and document processing, the recognition of handwritten numerals plays a crucial role. This work has led to remarkable changes in OCR, historical handwritten document analysis, and postal automation. In this study, we present a novel framework to overcome this challenge, going beyond digit recognition in only one language. Unlike common methods that focus on a limited set of languages, our method provides a comprehensive solution for recognition of handwritten digit images in 12 different languages. These specific languages are chosen because most of them have fairly distant representations in latent space. We utilize transfer learning, as it reduces the computational cost and maintains the quality of enhanced images and the models’ recognition accuracy. Another strength of our approach is the innovative attention-based module called the MRA module. Our experiments confirm that by applying this module, major progress is made in both image quality and the accuracy of handwritten digit recognition. Notably, we reached high precisions, surpassing nearly 2% improvement in specific languages compared to earlier techniques. In this work, we present a robust and cost-effective approach that handles multilingual handwritten numeral recognition across a wide range of languages. The code and further implementation details are available at https://github.com/CVLab-SHUT/HandWrittenDigitRecognition.
Emergency department's (ED) boarding (defined as ED waiting time greater than four hours) has been linked to poor patient outcomes and health system performance. Yet, effective forecasting models is rare before COVID-19, lacking during the peri-COVID era. Here, a hybrid convolutional neural network (CNN)-Long short-term memory (LSTM) model was applied to public-domain data sourced from Hong Kong's Hospital Authority, Department of Health, and Housing Authority. In addition, we sought to identify the phase of the COVID-19 pandemic that most significantly perturbed our complex adaptive healthcare system, thereby revealing a stable pattern of interconnectedness among its components, using deep transfer learning methodology. Our result shows that 1) the greatest proportion of days with ED boarding was found between waves four and five; 2) the best-performing model for forecasting ED boarding was observed between waves four and five, which was based on features representing time-invariant residential buildings' built environment and sociodemographic profiles and the historical time series of ED boarding and case counts, compared to during the waves when best-performing forecasting is based on time-series features alone; and 3) when the model built from the period between waves four and five was applied to data from other waves via deep transfer learning, the transferred model enhanced the performance of indigenous models.