Addressing the detrimental impact of non-stationary environmental noise on automatic speech recognition (ASR) has been a persistent and significant research focus. Despite advancements, this challenge continues to be a major concern. Recently, data-driven supervised approaches, such as deep neural networks, have emerged as promising alternatives to traditional unsupervised methods. With extensive training, these approaches have the potential to overcome the challenges posed by diverse real-life acoustic environments. In this light, this paper introduces a novel neural framework that incorporates a robust frontend into ASR systems in both clean and noisy environments. Utilizing the Aurora-2 speech database, the authors evaluate the effectiveness of an acoustic feature set for Mel-frequency, employing the approach of transfer learning based on Residual neural network (ResNet). The experimental results demonstrate a significant improvement in recognition accuracy compared to convolutional neural networks (CNN) and long short-term memory (LSTM) networks. They achieved accuracies of 98.94% in clean and 91.21% in noisy mode.
Human motion modeling traditionally separates motion generation and estimation into distinct tasks with specialized models. Motion generation models focus on creating diverse, realistic motions from inputs like text, audio, or keyframes, while motion estimation models aim to reconstruct accurate motion trajectories from observations like videos. Despite sharing underlying representations of temporal dynamics and kinematics, this separation limits knowledge transfer between tasks and requires maintaining separate models. We present GENMO, a unified Generalist Model for Human Motion that bridges motion estimation and generation in a single framework. Our key insight is to reformulate motion estimation as constrained motion generation, where the output motion must precisely satisfy observed conditioning signals. Leveraging the synergy between regression and diffusion, GENMO achieves accurate global motion estimation while enabling diverse motion generation. We also introduce an estimation-guided training objective that exploits in-the-wild videos with 2D annotations and text descriptions to enhance generative diversity. Furthermore, our novel architecture handles variable-length motions and mixed multimodal conditions (text, audio, video) at different time intervals, offering flexible control. This unified approach creates synergistic benefits: generative priors improve estimated motions under challenging conditions like occlusions, while diverse video data enhances generation capabilities. Extensive experiments demonstrate GENMO's effectiveness as a generalist framework that successfully handles multiple human motion tasks within a single model.
Large Language Models (LLM) are typically trained on vast amounts of data from various sources. Even when designed modularly (e.g., Mixture-of-Experts), LLMs can leak privacy on their sources. Conversely, training such models in isolation arguably prohibits generalization. To this end, we propose a framework, NoEsis, which builds upon the desired properties of modularity, privacy, and knowledge transfer. NoEsis integrates differential privacy with a hybrid two-staged parameter-efficient fine-tuning that combines domain-specific low-rank adapters, acting as experts, with common prompt tokens, acting as a knowledge-sharing backbone. Results from our evaluation on CodeXGLUE showcase that NoEsis can achieve provable privacy guarantees with tangible knowledge transfer across domains, and empirically show protection against Membership Inference Attacks. Finally, on code completion tasks, NoEsis bridges at least 77% of the accuracy gap between the non-shared and the non-private baseline.
This paper presents a comprehensive evaluation of lightweight deep learning models for image classification, emphasizing their suitability for deployment in resource-constrained environments such as low-memory devices. Five state-of-the-art architectures - MobileNetV3 Small, ResNet18, SqueezeNet, EfficientNetV2-S, and ShuffleNetV2 - are benchmarked across three diverse datasets: CIFAR-10, CIFAR-100, and Tiny ImageNet. The models are assessed using four key performance metrics: classification accuracy, inference time, floating-point operations (FLOPs), and model size. Additionally, we investigate the impact of hyperparameter tuning, data augmentation, and training paradigms by comparing pretrained models with scratch-trained counterparts, focusing on MobileNetV3 Small. Our findings reveal that transfer learning significantly enhances model accuracy and computational efficiency, particularly for complex datasets like Tiny ImageNet. EfficientNetV2 consistently achieves the highest accuracy, while MobileNetV3 offers the best balance between accuracy and efficiency, and SqueezeNet excels in inference speed and compactness. This study highlights critical trade-offs between accuracy and efficiency, offering actionable insights for deploying lightweight models in real-world applications where computational resources are limited. By addressing these challenges, this research contributes to optimizing deep learning systems for edge computing and mobile platforms.
Physics-Informed Neural Networks have emerged as a promising methodology for solving PDEs, gaining significant attention in computer science and various physics-related fields. Despite being demonstrated the ability to incorporate the physics of laws for versatile applications, PINNs still struggle with the challenging problems which are stiff to be solved and/or have high-frequency components in the solutions, resulting in accuracy and convergence issues. It may not only increase computational costs, but also lead to accuracy loss or solution divergence. In this study, an alternative approach is proposed to mitigate the above-mentioned problems. Inspired by the multi-grid method in CFD community, the underlying idea of the current approach is to efficiently remove different frequency errors via training with different levels of training samples, resulting in a simpler way to improve the training accuracy without spending time in fine-tuning of neural network structures, loss weights as well as hyperparameters. To demonstrate the efficacy of current approach, we first investigate canonical 1D ODE with high-frequency component and 2D convection-diffusion equation with V-cycle training strategy. Finally, the current method is employed for the classical benchmark problem of steady Lid-driven cavity flows at different Reynolds numbers, to investigate the applicability and efficacy for the problem involved multiple modes of high and low frequency. By virtue of various training sequence modes, improvement through predictions lead to 30% to 60% accuracy improvement. We also investigate the synergies between current method and transfer learning techniques for more challenging problems (i.e., higher Re). From the present results, it also revealed that the current framework can produce good predictions even for the case of Re=5000, demonstrating the ability to solve complex high-frequency PDEs.
The deployment of machine learning (ML)-based process monitoring systems has significantly advanced additive manufacturing (AM) by enabling real-time defect detection, quality assessment, and process optimization. However, redundancy is a critical yet often overlooked challenge in the deployment and operation of ML-based AM process monitoring systems. Excessive redundancy leads to increased equipment costs, compromised model performance, and high computational requirements, posing barriers to industrial adoption. However, existing research lacks a unified definition of redundancy and a systematic framework for its evaluation and mitigation. This paper defines redundancy in ML-based AM process monitoring and categorizes it into sample-level, feature-level, and model-level redundancy. A comprehensive multi-level redundancy mitigation (MLRM) framework is proposed, incorporating advanced methods such as data registration, downscaling, cross-modality knowledge transfer, and model pruning to systematically reduce redundancy while improving model performance. The framework is validated through an ML-based in-situ defect detection case study for directed energy deposition (DED), demonstrating a 91% reduction in latency, a 47% decrease in error rate, and a 99.4% reduction in storage requirements. Additionally, the proposed approach lowers sensor costs and energy consumption, enabling a lightweight, cost-effective, and scalable monitoring system. By defining redundancy and introducing a structured mitigation framework, this study establishes redundancy analysis and mitigation as a key enabler of efficient ML-based process monitoring in production environments.
Semi-supervised learning has become a compelling approach for 3D tooth segmentation from CBCT scans, where labeled data is minimal. However, existing methods still face two persistent challenges: limited corrective supervision in structurally ambiguous or mislabeled regions during supervised training and performance degradation caused by unreliable pseudo-labels on unlabeled data. To address these problems, we propose Region-Aware Instructive Learning (RAIL), a dual-group dual-student, semi-supervised framework. Each group contains two student models guided by a shared teacher network. By alternating training between the two groups, RAIL promotes intergroup knowledge transfer and collaborative region-aware instruction while reducing overfitting to the characteristics of any single model. Specifically, RAIL introduces two instructive mechanisms. Disagreement-Focused Supervision (DFS) Controller improves supervised learning by instructing predictions only within areas where student outputs diverge from both ground truth and the best student, thereby concentrating supervision on structurally ambiguous or mislabeled areas. In the unsupervised phase, Confidence-Aware Learning (CAL) Modulator reinforces agreement in regions with high model certainty while reducing the effect of low-confidence predictions during training. This helps prevent our model from learning unstable patterns and improves the overall reliability of pseudo-labels. Extensive experiments on four CBCT tooth segmentation datasets show that RAIL surpasses state-of-the-art methods under limited annotation. Our code will be available at https://github.com/Tournesol-Saturday/RAIL.
Sepsis is a life threatening medical condition that occurs when the body has an extreme response to infection, leading to widespread inflammation, organ failure, and potentially death. Because sepsis can worsen rapidly, early detection is critical to saving lives. However, current diagnostic methods often identify sepsis only after significant damage has already occurred. Our project aims to address this challenge by developing a machine learning based system to predict sepsis in its early stages, giving healthcare providers more time to intervene. A major problem with existing models is the wide variability in the patient information or features they use, such as heart rate, temperature, and lab results. This inconsistency makes models difficult to compare and limits their ability to work across different hospitals and settings. To solve this, we propose a method called Feature Aligned Transfer Learning (FATL), which identifies and focuses on the most important and commonly reported features across multiple studies, ensuring the model remains consistent and clinically relevant. Most existing models are trained on narrow patient groups, leading to population bias. FATL addresses this by combining knowledge from models trained on diverse populations, using a weighted approach that reflects each models contribution. This makes the system more generalizable and effective across different patient demographics and clinical environments. FATL offers a practical and scalable solution for early sepsis detection, particularly in hospitals with limited resources, and has the potential to improve patient outcomes, reduce healthcare costs, and support more equitable healthcare delivery.
This paper explores the use of an extended neuroevolutionary approach, based on NeuroEvolution of Augmenting Topologies (NEAT), for autonomous robots in dynamic environments associated with hazardous tasks like firefighting, urban search-and-rescue (USAR), and industrial inspections. Building on previous research, it expands the simulation environment to larger and more complex settings, demonstrating NEAT's adaptability across different applications. By integrating recent advancements in NEAT and reinforcement learning, the study uses modern simulation frameworks for realism and hybrid algorithms for optimization. Experimental results show that NEAT-evolved controllers achieve success rates comparable to state-of-the-art deep reinforcement learning methods, with superior structural adaptability. The agents reached ~80% success in outdoor tests, surpassing baseline models. The paper also highlights the benefits of transfer learning among tasks and evaluates the effectiveness of NEAT in complex 3D navigation. Contributions include evaluating NEAT for diverse autonomous applications and discussing real-world deployment considerations, emphasizing the approach's potential as an alternative or complement to deep reinforcement learning in autonomous navigation tasks.
Knowledge distillation is a widely used technique for compressing large language models (LLMs) by training a smaller student model to mimic a larger teacher model. Typically, both the teacher and student are Transformer-based architectures, leveraging softmax attention for sequence modeling. However, the quadratic complexity of self-attention at inference time remains a significant bottleneck, motivating the exploration of subquadratic alternatives such as structured state-space models (SSMs), linear attention, and recurrent architectures. In this work, we systematically evaluate the transferability of knowledge distillation from a Transformer teacher to nine subquadratic student architectures. Our study aims to determine which subquadratic model best aligns with the teacher's learned representations and how different architectural constraints influence the distillation process. We also investigate the impact of intelligent initialization strategies, including matrix mixing and query-key-value (QKV) copying, on the adaptation process. Our empirical results on multiple NLP benchmarks provide insights into the trade-offs between efficiency and performance, highlighting key factors for successful knowledge transfer to subquadratic architectures.
Parameter-efficient transfer learning (PETL) methods adapt large artificial neural networks to downstream tasks without fine-tuning the entire model. However, existing additive methods, such as adapters, sometimes struggle to capture distributional shifts in intermediate feature embeddings. We propose a novel histogram-based parameter-efficient tuning (HPT) technique that captures the statistics of the target domain and modulates the embeddings. Experimental results on three downstream passive sonar datasets (ShipsEar, DeepShip, VTUAD) demonstrate that HPT outperforms conventional adapters. Notably, HPT achieves 91.8% vs. 89.8% accuracy on VTUAD. Furthermore, HPT trains faster and yields feature representations closer to those of fully fine-tuned models. Overall, HPT balances parameter savings and performance, providing a distribution-aware alternative to existing adapters and shows a promising direction for scalable transfer learning in resource-constrained environments. The code is publicly available: https://github.com/Advanced-Vision-and-Learning-Lab/HLAST_DeepShip_ParameterEfficient.
Making it intelligent is a promising way in System/OS design. This paper proposes OSML+, a new ML-based resource scheduling mechanism for co-located cloud services. OSML+ intelligently schedules the cache and main memory bandwidth resources at the memory hierarchy and the computing core resources simultaneously. OSML+ uses a multi-model collaborative learning approach during its scheduling and thus can handle complicated cases, e.g., avoiding resource cliffs, sharing resources among applications, enabling different scheduling policies for applications with different priorities, etc. OSML+ can converge faster using ML models than previous studies. Moreover, OSML+ can automatically learn on the fly and handle dynamically changing workloads accordingly. Using transfer learning technologies, we show our design can work well across various cloud servers, including the latest off-the-shelf large-scale servers. Our experimental results show that OSML+ supports higher loads and meets QoS targets with lower overheads than previous studies.
Demand for 2K video synthesis is rising with increasing consumer expectations for ultra-clear visuals. While diffusion transformers (DiTs) have demonstrated remarkable capabilities in high-quality video generation, scaling them to 2K resolution remains computationally prohibitive due to quadratic growth in memory and processing costs. In this work, we propose Turbo2K, an efficient and practical framework for generating detail-rich 2K videos while significantly improving training and inference efficiency. First, Turbo2K operates in a highly compressed latent space, reducing computational complexity and memory footprint, making high-resolution video synthesis feasible. However, the high compression ratio of the VAE and limited model size impose constraints on generative quality. To mitigate this, we introduce a knowledge distillation strategy that enables a smaller student model to inherit the generative capacity of a larger, more powerful teacher model. Our analysis reveals that, despite differences in latent spaces and architectures, DiTs exhibit structural similarities in their internal representations, facilitating effective knowledge transfer. Second, we design a hierarchical two-stage synthesis framework that first generates multi-level feature at lower resolutions before guiding high-resolution video generation. This approach ensures structural coherence and fine-grained detail refinement while eliminating redundant encoding-decoding overhead, further enhancing computational efficiency.Turbo2K achieves state-of-the-art efficiency, generating 5-second, 24fps, 2K videos with significantly reduced computational cost. Compared to existing methods, Turbo2K is up to 20$\times$ faster for inference, making high-resolution video generation more scalable and practical for real-world applications.
As machine learning evolves, domain generalization (DG) and domain adaptation (DA) have become crucial for enhancing model robustness across diverse environments. Contrastive Language-Image Pretraining (CLIP) plays a significant role in these tasks, offering powerful zero-shot capabilities that allow models to perform effectively in unseen domains. However, there remains a significant gap in the literature, as no comprehensive survey currently exists that systematically explores the applications of CLIP in DG and DA, highlighting the necessity for this review. This survey presents a comprehensive review of CLIP's applications in DG and DA. In DG, we categorize methods into optimizing prompt learning for task alignment and leveraging CLIP as a backbone for effective feature extraction, both enhancing model adaptability. For DA, we examine both source-available methods utilizing labeled source data and source-free approaches primarily based on target domain data, emphasizing knowledge transfer mechanisms and strategies for improved performance across diverse contexts. Key challenges, including overfitting, domain diversity, and computational efficiency, are addressed, alongside future research opportunities to advance robustness and efficiency in practical applications. By synthesizing existing literature and pinpointing critical gaps, this survey provides valuable insights for researchers and practitioners, proposing directions for effectively leveraging CLIP to enhance methodologies in domain generalization and adaptation. Ultimately, this work aims to foster innovation and collaboration in the quest for more resilient machine learning models that can perform reliably across diverse real-world scenarios. A more up-to-date version of the papers is maintained at: https://github.com/jindongli-Ai/Survey_on_CLIP-Powered_Domain_Generalization_and_Adaptation.
Deep learning algorithms have significantly reduced the computational time and improved the spatial resolution of particle image velocimetry~(PIV). However, the models trained on synthetic datasets might have a degraded performance on practical particle images due to domain gaps. As a result, special residual patterns are often observed for the vector fields of deep learning-based estimators. To reduce the special noise step-by-step, we employ a denoising diffusion model~(FlowDiffuser) for PIV analysis. And the data-hungry iterative denoising diffusion model is trained via a transfer learning strategy, resulting in our PIV-FlowDiffuser method. Specifically, (1) pre-training a FlowDiffuser model with multiple optical flow datasets of the computer vision community, such as Sintel, KITTI, etc; (2) fine-tuning the pre-trained model on synthetic PIV datasets. Note that the PIV images are upsampled by a factor of two to resolve the small-scale turbulent flow structures. The visualized results indicate that our PIV-FlowDiffuser effectively suppresses the noise patterns. Therefore, the denoising diffusion model reduces the average end-point error~($AEE$) by 59.4% over RAFT256-PIV baseline on the classic Cai's dataset. Besides, PIV-FlowDiffuser exhibits enhanced generalization performance on unseen particle images due to transfer learning. Overall, this study highlights the transfer-learning-based denoising diffusion models for PIV. And a detailed implementation is recommended for interested readers in the repository https://github.com/Zhu-Qianyu/PIV-FlowDiffuser.
In practical scenarios, processes such as sensor design, manufacturing, and installation will introduce certain errors. Furthermore, mutual interference occurs when the sensors receive signals. These defects in array systems are referred to as array imperfections, which can significantly degrade the performance of Direction of Arrival (DOA) estimation. In this study, we propose a deep-learning based transfer learning approach, which effectively mitigates the degradation of deep-learning based DOA estimation performance caused by array imperfections. In the proposed approach, we highlight three major contributions. First, we propose a Vision Transformer (ViT) based method for DOA estimation, which achieves excellent performance in scenarios with low signal-to-noise ratios (SNR) and limited snapshots. Second, we introduce a transfer learning framework that extends deep learning models from ideal simulation scenarios to complex real-world scenarios with array imperfections. By leveraging prior knowledge from ideal simulation data, the proposed transfer learning framework significantly improves deep learning-based DOA estimation performance in the presence of array imperfections, without the need for extensive real-world data. Finally, we incorporate visualization and evaluation metrics to assess the performance of DOA estimation algorithms, which allow for a more thorough evaluation of algorithms and further validate the proposed method. Our code can be accessed at https://github.com/zzb-nice/DOA_est_Master.
The rapid advancement of deep learning (DL) has transformed healthcare, particularly in cancer detection and diagnosis. DL surpasses traditional machine learning and human accuracy, making it a critical tool for identifying diseases. Despite numerous reviews on DL in healthcare, a comprehensive analysis of its role in cancer detection remains limited. Existing studies focus on specific aspects, leaving gaps in understanding its broader impact. This paper addresses these gaps by reviewing advanced DL techniques, including transfer learning (TL), reinforcement learning (RL), federated learning (FL), Transformers, and large language models (LLMs). These approaches enhance accuracy, tackle data scarcity, and enable decentralized learning while maintaining data privacy. TL adapts pre-trained models to new datasets, improving performance with limited labeled data. RL optimizes diagnostic pathways and treatment strategies, while FL fosters collaborative model development without sharing sensitive data. Transformers and LLMs, traditionally used in natural language processing, are now applied to medical data for improved interpretability. Additionally, this review examines these techniques' efficiency in cancer diagnosis, addresses challenges like data imbalance, and proposes solutions. It serves as a resource for researchers and practitioners, providing insights into current trends and guiding future research in advanced DL for cancer detection.
This paper explores transfer learning in heterogeneous multi-source environments with distributional divergence between target and auxiliary domains. To address challenges in statistical bias and computational efficiency, we propose a Sparse Optimization for Transfer Learning (SOTL) framework based on L0-regularization. The method extends the Joint Estimation Transferred from Strata (JETS) paradigm with two key innovations: (1) L0-constrained exact sparsity for parameter space compression and complexity reduction, and (2) refining optimization focus to emphasize target parameters over redundant ones. Simulations show that SOTL significantly improves both estimation accuracy and computational speed, especially under adversarial auxiliary domain conditions. Empirical validation on the Community and Crime benchmarks demonstrates the statistical robustness of the SOTL method in cross-domain transfer.
This paper develops a new spectral clustering-based method called TransNet for transfer learning in community detection of network data. Our goal is to improve the clustering performance of the target network using auxiliary source networks, which are heterogeneous, privacy-preserved, and locally stored across various sources. The edges of each locally stored network are perturbed using the randomized response mechanism to achieve differential privacy. Notably, we allow the source networks to have distinct privacy-preserving and heterogeneity levels as often desired in practice. To better utilize the information from the source networks, we propose a novel adaptive weighting method to aggregate the eigenspaces of the source networks multiplied by adaptive weights chosen to incorporate the effects of privacy and heterogeneity. We propose a regularization method that combines the weighted average eigenspace of the source networks with the eigenspace of the target network to achieve an optimal balance between them. Theoretically, we show that the adaptive weighting method enjoys the error-bound-oracle property in the sense that the error bound of the estimated eigenspace only depends on informative source networks. We also demonstrate that TransNet performs better than the estimator using only the target network and the estimator using only the weighted source networks.
Sample efficiency is critical for online Reinforcement Learning from Human Feedback (RLHF). While existing works investigate sample-efficient online exploration strategies, the potential of utilizing misspecified yet relevant reward models to accelerate learning remains underexplored. This paper studies how to transfer knowledge from those imperfect reward models in online RLHF. We start by identifying a novel property due to KL-regularization in the RLHF objective: \emph{a policy's coverability of the optimal policy is captured by its sub-optimality}. Building on this insight, we propose novel transfer learning principles and a theoretical algorithm -- \emph{\textbf{T}ransfer \textbf{P}olicy \textbf{O}ptimization (\textbf{TPO})} -- with provable benefits compared to standard online learning. Empirically, inspired by our theoretical findings, we develop a win-rate-based transfer policy selection strategy with improved computational efficiency. Moreover, our empirical transfer learning technique is modular and can be integrated with various policy optimization methods, such as DPO, IPO and XPO, to further enhance their performance. We validate the effectiveness of our method through experiments on summarization tasks.