Visual reinforcement learning agents typically face serious performance declines in real-world applications caused by visual distractions. Existing methods rely on fine-tuning the policy's representations with hand-crafted augmentations. In this work, we propose Self-Consistent Model-based Adaptation (SCMA), a novel method that fosters robust adaptation without modifying the policy. By transferring cluttered observations to clean ones with a denoising model, SCMA can mitigate distractions for various policies as a plug-and-play enhancement. To optimize the denoising model in an unsupervised manner, we derive an unsupervised distribution matching objective with a theoretical analysis of its optimality. We further present a practical algorithm to optimize the objective by estimating the distribution of clean observations with a pre-trained world model. Extensive experiments on multiple visual generalization benchmarks and real robot data demonstrate that SCMA effectively boosts performance across various distractions and exhibits better sample efficiency.
论文检索
输入标题、作者或关键词，从 1,280 篇学术成果中精准定位
Spurious correlations pose a significant challenge to the robustness of statistical models, often resulting in unsatisfactory performance when distributional shifts occur between training and testing data. To address this, we propose to transfer knowledge across spuriously correlated categories within the deep feature space. Specifically, samples' deep features are enriched using semantic vectors extracted from both their respective category distributions and those of their spuriously correlated counterparts, enabling the generation of diverse class-specific factual and counterfactual augmented deep features. We then demonstrate the feasibility of optimizing a surrogate robust loss instead of conducting explicit augmentations by considering an infinite number of augmentations. As spurious correlations between samples and classes evolve during training, we develop a reinforcement learning-based training framework called Dynamic Knowledge Transfer (DKT) to facilitate dynamic adjustments in the direction and intensity of knowledge transfer. Within this framework, a target network is trained using the derived robust loss to enhance robustness, while a strategy network generates sample-wise augmentation strategies in a dynamic and automatic way. Extensive experiments validate the effectiveness of the DKT framework in mitigating spurious correlations, achieving state-of-the-art performance across three typical learning scenarios susceptible to such correlations.
Enhancing the Performance of Global Model by Improving the Adaptability of Local Models in Federated Learning
PDF ↗Federated learning enables the clients to collaboratively train a global model, which is aggregated from local models. Due to the heterogeneous data distributions over clients and data privacy in federated learning, it is difficult to train local models to achieve a well-performed global model. In this paper, we introduce the adaptability of local models, i.e., the average performance of local models on data distributions over clients, and enhance the performance of the global model by improving the adaptability of local models. Since each client does not know the data distributions over other clients, the adaptability of the local model cannot be directly optimized. First, we provide the property of an appropriate local model which has good adaptability on the data distributions over clients. Then, we formalize the property into the local training objective with a constraint and propose a feasible solution to train the local model. Extensive experiments on federated learning benchmarks demonstrate that our method significantly improves the adaptability of local models and achieves a well-performed global model that consistently outperforms the baseline methods.
Physics-informed neural networks (PINNs) have emerged as a new learning paradigm for solving partial differential equations (PDEs) by enforcing the constraints of physical equations, boundary conditions (BCs), and initial conditions (ICs) into the loss function. Despite their successes, vanilla PINNs still suffer from poor accuracy and slow convergence due to the intractable multi-objective optimization issue. In this paper, we propose a novel Dual-Balanced PINN (DB-PINN), which dynamically adjusts loss weights by integrating inter-balancing and intra-balancing to alleviate two imbalance issues in PINNs. Inter-balancing aims to mitigate the gradient imbalance between PDE residual loss and condition-fitting losses by determining an aggregated weight that offsets their gradient distribution discrepancies. Intra-balancing acts on condition-fitting losses to tackle the imbalance in fitting difficulty across diverse conditions. By evaluating the fitting difficulty based on the loss records, intra-balancing can allocate the aggregated weight proportionally to each condition loss according to its fitting difficulty level. We further introduce a robust weight update strategy to prevent abrupt spikes and arithmetic overflow in instantaneous weight values caused by large loss variances, enabling smooth weight updating and stable training. Extensive experiments demonstrate that DB-PINN achieves significantly superior performance than those popular gradient-based weighting methods in terms of convergence speed and prediction accuracy. Our code and supplementary material are available at https://github.com/chenhong-zhou/DualBalanced-PINNs.
Spatio-temporal time series (STTS) have been widely used in many applications. However, accurately forecasting STTS is challenging due to complex dynamic correlations in both time and space dimensions. Existing graph neural networks struggle to balance effectiveness and efficiency in modeling dynamic spatio-temporal relations. To address this problem, we propose the Dynamic Spatio-Temporal Stiefel Graph Neural Network (DST-SGNN) to efficiently process STTS. For DST-SGNN, we first introduce the novel Stiefel Graph Spectral Convolution (SGSC) and Stiefel Graph Fourier Transform (SGFT). The SGFT matrix in SGSC is constrained to lie on the Stiefel manifold, and SGSC can be regarded as a filtered graph spectral convolution. We also propose the Linear Dynamic Graph Optimization on Stiefel Manifold (LDGOSM), which can efficiently learn the SGFT matrix from the dynamic graph and significantly reduce the computational complexity. Finally, we propose a multi-layer SGSC (MSGSC) that efficiently captures complex spatio-temporal correlations. Extensive experiments on seven spatio-temporal datasets show that DST-SGNN outperforms state-of-the-art methods while maintaining relatively low computational costs.
Deep learning models often achieve high performance by inadvertently learning spurious correlations between targets and non-essential features. For example, an image classifier may identify an object via its background that spuriously correlates with it. This prediction behavior, known as spurious bias, severely degrades model performance on data that lacks the learned spurious correlations. Existing methods on spurious bias mitigation typically require a variety of data groups with spurious correlation annotations called group labels. However, group labels require costly human annotations and often fail to capture subtle spurious biases such as relying on specific pixels for predictions. In this paper, we propose a novel post hoc spurious bias mitigation framework without requiring group labels. Our framework, termed ShortcutProbe, identifies prediction shortcuts that reflect potential non-robustness in predictions in a given model's latent space. The model is then retrained to be invariant to the identified prediction shortcuts for improved robustness. We theoretically analyze the effectiveness of the framework and empirically demonstrate that it is an efficient and practical tool for improving a model's robustness to spurious bias on diverse datasets.
Graph contrastive learning (GCL) often suffers from false negatives, which degrades the performance on downstream tasks. The existing methods addressing the false negative issue usually rely on human prior knowledge, still leading GCL to suboptimal results. In this paper, we propose a novel Negative Metric Learning (NML) enhanced GCL (NML-GCL). NML-GCL employs a learnable Negative Metric Network (NMN) to build a negative metric space, in which false negatives can be distinguished better from true negatives based on their distance to anchor node. To overcome the lack of explicit supervision signals for NML, we propose a joint training scheme with bi-level optimization objective, which implicitly utilizes the self-supervision signals to iteratively optimize the encoder and the negative metric network. The solid theoretical analysis and the extensive experiments conducted on widely used benchmarks verify the superiority of the proposed method.
Predicting the performance of deep learning (DL) models, such as execution time and resource utilization, is crucial for Neural Architecture Search (NAS), DL cluster schedulers, and other technologies that advance deep learning. The representation of a model is the foundation for its performance prediction. However, existing methods cannot comprehensively represent diverse model configurations, resulting in unsatisfactory accuracy. To address this, we represent a model as a graph that includes the topology, along with node, edge, and global features, all of which are crucial for effectively capturing the performance of the model. Based on this representation, we propose PerfSeer, a novel predictor that uses a Graph Neural Network (GNN)-based performance prediction model, SeerNet. SeerNet fully leverages the topology and various features, while incorporating optimizations such as Synergistic Max-Mean aggregation (SynMM) and Global-Node Perspective Boost (GNPB) to more effectively capture the critical performance information, enabling it to predict the performance of models accurately. Furthermore, SeerNet can be extended to SeerNet-Multi by using Project Conflicting Gradients (PCGrad), enabling efficient simultaneous prediction of multiple performance metrics without significantly affecting accuracy. We constructed a dataset containing performance metrics for 53k+ model configurations, including execution time, memory usage, and Streaming Multiprocessor (SM) utilization during both training and inference. The evaluation results show that PerfSeer outperforms nn-Meter, Brp-NAS, and DIPPM.
High-Confident Local Structure Guided Consensus Graph Learning For Incomplete Multi-view Clustering
PDF ↗Current existing clustering methods for handling incomplete multi-view data primarily concentrate on learning a common representation or graph from the available views, while overlooking the latent information contained in the missing views and the imbalance of information among different views. Furthermore, instances with weak discriminative features usually degrading the precision of consistent representation or graph across all views. To address these problems, in this paper, we propose a simple but efficient method, called high-confident local structure guided consensus graph learning for incomplete multi-view clustering (HLSCG_IMC). Specifically, this method can adaptively learn a strict block diagonal structure from the available samples using a block diagonal representation regularizer. Different from the existing methods using a simple pairwise affinity graph for structure construction, we consider the influence of instances located at the edge of two clusters on the construction of graph for each view. By harnessing the proposed high-confident strict block diagonal structures, the approach seeks to directly guide the learning of the robust consensus graph. A number of experiments have been conducted to verify the efficacy of our approach.
Instance Relation Learning Network with Label Knowledge Propagation for Few-shot Multi-label Intent Detection
PDF ↗Few-shot Multi-label Intent Detection (MID) is crucial for dialogue systems, aiming to detect multiple intents of utterances in low-resource dialogue domains. Previous studies focus on a two-stage pipeline. They first learn representations of utterances with multiple labels and then use a threshold-based strategy to identify multi-label results. However, these methods rely on representation classification and ignore instance relations, leading to error propagation. To solve the above issues, we propose a multi-label joint learning method for few-shot MID in an end-to-end manner, which constructs an instance relation learning network with label knowledge propagation to eliminate error propagation. Concretely, we learn the interaction relations between instances with class information to propagate label knowledge between a few labeled (support set) and unlabeled (query set) instances. With label knowledge propagation, the relation strength between instances directly indicates whether two utterances belong to the same intent for multi-label prediction. Besides, a dual relation-enhanced loss is developed to optimize support- and query-level relation strength to improve performance. Experiments show that we outperform strong baselines by an average of 9.54% AUC and 11.19% Macro-F1 in 1-shot scenarios.
DM-POSA: Enhancing Open-World Test-Time Adaptation with Dual-Mode Matching and Prompt-Based Open Set Adaptation
PDF ↗The need to generalize the pre-trained deep learning models to unknown test-time data distributions has spurred research into test-time adaptation (TTA). Existing studies have mainly focused on closed-set TTA with only covariate shifts, while largely overlooking open-set TTA that involves semantic shifts, i.e., unknown open-set classes. However, addressing adaptation to unknown classes is crucial for open-world safety-critical applications such as autonomous driving. In this paper, we emphasize that accurate identification of the open-set samples is rather challenging in TTA. The entanglement of semantic shift and covariate shift mutually confuse the network’s discriminative capability. This co-interference further exacerbates considering the single-pass data nature and low latency requirements. With this under standing, we propose Dual-mode Matching and Prompt-based Open Set Adaptation (DM-POSA) for open-set TTA to enhance discriminative feature learning and unknown classes distinguishment with minimal time cost. DM-POSA identifies open-set samples via dual-mode matching strategies, including model-parameter-based and feature space-based matching. It also optimizes the model with a random pairing discrepancy loss, enhancing the distributional difference between open-set and closed-set samples, thus improving the model’s ability to recognize unknown categories. Extensive experiments show the superiority of DM-POSA over state-of-the-art baselines on both closed-set class adaptation and open-set class detection.
Recently, multi-view data has gradually attracted attention. However, real-world applications often face Partial View-aligned Problem (PVP) and Partially Sample-missing Problem (PSP) due to data loss or corruption. Existing methods addressing PVP typically focus only on learning from the information of aligned data, while ignoring unaligned data where samples exist but lack alignment relationships. This introduces PSP, which does not inherently exist in the data, leading to biased learning of the data's information. For PSP, due to varying degrees of missing data, incomplete spatial structures can cause clustering centers-shifted problem, resulting in the model learning incorrect correspondences and biased spatial structures.To tackle them, we propose a novel method called Dual Robust Unbiased Multi-View Clustering for Incomplete and Unpaired Information (DRUMVC). To our knowledge, this is the first noise-robust and unbiased multi-view clustering method capable of simultaneously addressing both PVP and PSP. Specifically, DRUMVC leverages aligned and complete samples as a bridge to construct high-quality correspondences for samples lacking cross-view relationship information due to PVP or PSP. Additionally, we employ a dual noise-robust contrastive learning loss to mitigate the impact of noise potentially introduced during the pair construction. Experiments on several challenging datasets demonstrate the superiority of our proposed method.
Experience replay is widely used to improve learning efficiency in reinforcement learning by leveraging past experiences. However, existing experience replay methods, whether based on uniform or prioritized sampling, often suffer from low efficiency, particularly in real-world scenarios with high-dimensional state spaces. To address this limitation, we propose a novel approach, Efficient Diversity-based Experience Replay (EDER). EDER employs a determinantal point process to model the diversity between samples and prioritizes replay based on the diversity between samples. To further enhance learning efficiency, we incorporate Cholesky decomposition for handling large state spaces in realistic environments. Additionally, rejection sampling is applied to select samples with higher diversity, thereby improving overall learning efficacy. Extensive experiments are conducted on robotic manipulation tasks in MuJoCo, Atari games, and realistic indoor environments in Habitat. The results demonstrate that our approach not only significantly improves learning efficiency but also achieves superior performance in high-dimensional, realistic environments.
AdaptPFL: Unlocking Cross-Device Palmprint Recognition via Adaptive Personalized Federated Learning with Feature Decoupling
PDF ↗Contactless palmprint recognition has recently emerged as a promising biometric technology. However, traditional methods that require sharing user data introduce substantial security risks. While federated learning offers privacy-preserving solutions, it often compromises recognition accuracy due to feature distribution drift caused by external factors such as lighting and devices. To address this issue, we propose an adaptive personalized federated learning framework (AdaptPFL). The central innovation lies in decomposing palmprint features into identity-related and contextual-related components using a feature decoupling mechanism. This design isolates the influence of external environmental factors on identity recognition through de-entanglement. Furthermore, two adaptive aggregation strategies are introduced to correct client drift: (1) Intra-Local Adaptive Aggregation (ILAA), which addresses intra-client drift by adaptively combining the two decoupled feature types; (2) Global-Local Adaptive Aggregation (GLAA), which corrects inter-client drift by adaptively aggregating model parameters. Experimental results demonstrate that AdaptPFL achieves superior performance compared to existing state-of-the-art methods.
Existing graph neural network (GNN) methods are typically built upon the i.i.d. assumption, emphasizing the enhancement of the test performance for in-distribution (ID) data. However, there has been limited exploration of their adaptability to scenarios involving unknown distribution data. On the one hand, in real-world application scenarios, graph data often expands continuously with the acquisition of external knowledge, which means that new nodes with unknown categories may be added to the graph data. The gap between the new node distribution and the original node distribution can make existing GNN methods less effective. On the other hand, existing out-of-distribution (OOD) detection methods often rely on the softmax confidence score, which makes the OOD data suffer from overconfident posterior distributions. To address the above issues, we propose an Energy Propagation-based Graph Neural Network (EPGNN), which improves the OOD generalization ability by endowing GNN with the capacity to detect the OOD nodes in the graph. Specifically, we first construct GNN encoder to obtain node embedding that incorporates neighborhood structural information. Then, we design a plug-and-play energy-based OOD evaluator by assigning corresponding energy values to different nodes. Finally, we construct a plug-and-play structure-aware energy propagation module and joint alignment regularization, which make the node energy more flexible during the training process. Extensive experiments on benchmark datasets demonstrate the superiority of our method.
The past decades witness the significant advancements in time series forecasting (TSF) across various real-world domains, including e-commerce and disease spread prediction. However, TSF is usually constrained by the uncertainty dilemma of predicting future data with limited past observations. To settle this question, we explore the use of Cross-Future Behavior (CFB) in TSF, which occurs before the current time but takes effect in the future. We leverage CFB features and propose the CRoss-Future Behavior Awareness based Time Series Forecasting method (CRAFT). The core idea of CRAFT is to utilize the trend of cross-future behavior to mine the trend of time series data to be predicted. Specifically, to settle the sparse and partial flaws of cross-future behavior, CRAFT employs the Koopman Predictor Module to extract the key trend and the Internal Trend Mining Module to supplement the unknown area of the cross-future behavior matrix. Then, we introduce the External Trend Guide Module with a hierarchical structure to acquire more representative trends from higher levels. Finally, we apply the demand-constrained loss to calibrate the distribution deviation of prediction results. We conduct experiments on real-world dataset. Experiments on both offline large-scale dataset and online A/B test demonstrate the effectiveness of CRAFT. Our dataset and code are available at https://github.com/CRAFTinTSF/CRAFT.
Federated stochastic bilevel optimization has been actively studied in recent years due to its widespread applications in machine learning. However, most existing federated stochastic bilevel optimization algorithms require the computation of second-order Hessian and Jacobian matrices, which leads to longer running times in practice. To address these challenges, we propose a novel federated stochastic variance-reduced bilevel gradient descent algorithm that relies solely on first-order oracles. Specifically, our approach does not require the computation of second-order Hessian and Jacobian matrices, significantly reducing running time. Furthermore, we introduce a novel learning rate mechanism, i.e., a constant single-time-scale learning rate, to coordinate the update of different variables. We also present a new strategy to establish the convergence rate of our algorithm. Finally, the extensive experimental results confirm the efficacy of our proposed algorithm.
QiMeng-TensorOp: One-Line Prompt is Enough for High-Performance Tensor Operator Generation with Hardware Primitives
PDF ↗Computation-intensive tensor operators constitute over 90% of the computations in Large Language Models (LLMs) and Deep Neural Networks. Automatically and efficiently generating high-performance tensor operators with hardware primitives is crucial for diverse and ever-evolving hardware architectures like RISC-V, ARM, and GPUs, as manually optimized implementation takes at least months and lacks portability. LLMs excel at generating high-level language codes, but they struggle to fully comprehend hardware characteristics and produce high-performance tensor operators. We introduce a tensor-operator auto-generation framework with a one-line user prompt (QiMeng-TensorOp), which enables LLMs to automatically exploit hardware characteristics to generate tensor operators with hardware primitives, and tune parameters for optimal performance across diverse hardware. Experimental results on various hardware platforms, SOTA LLMs, and typical tensor operators demonstrate that QiMeng-TensorOp effectively unleashes the computing capability of various hardware platforms, and automatically generates tensor operators of superior performance. Compared with vanilla LLMs, QiMeng-TensorOp achieves up to 1291× performance improvement. Even compared with human experts, QiMeng-TensorOp could reach 251% of OpenBLAS on RISC-V CPUs, and 124% of cuBLAS on NVIDIA GPUs. Additionally, QiMeng-TensorOp also significantly reduces development costs by 200× compared with human experts.
The representation of feature space is a crucial environment where data points get vectorized and embedded for subsequent modeling. Thus, the efficacy of machine learning (ML) algorithms is closely related to the quality of feature engineering. As one of the most important techniques, feature generation transforms raw data into an optimized feature space conducive to model training and further refines the space. Despite the advancements in automated feature engineering and feature generation, current methodologies often suffer from three fundamental issues: lack of explainability, limited applicability, and inflexible strategy. These shortcomings frequently hinder and limit the deployment of ML models across varied scenarios. Our research introduces a novel approach adopting large language models (LLMs) and feature-generating prompts to address these challenges. We propose a dynamic and adaptive feature generation method that enhances the interpretability of the feature generation process. Our approach broadens the applicability across various data types and tasks and offers advantages in terms of strategic flexibility. A broad range of experiments showcases that our approach is significantly superior to existing methods.
Humans excel in analogical learning and knowledge transfer and, more importantly, possess a unique understanding of identifying appropriate sources of knowledge. From a model's perspective, this presents a unique challenge. If models could autonomously retrieve knowledge relevant for transfer or decision-making to solve problems, they would transition from passively acquiring to actively accessing and learning from knowledge. However, filling models with knowledge is relatively straightforward—it simply requires more training and accessible knowledge bases. The more complex task is teaching models about which knowledge can be analogized and transferred. Therefore, we design a knowledge augmentation method, LEKA, for knowledge transfer that actively searches for suitable knowledge sources that can enrich the target domain's knowledge. This LEKA method extracts key information from the target domain's textual information, retrieves pertinent data from external data libraries, and harmonizes retrieved data with the target domain data in feature space and marginal probability measures. We validate the effectiveness of our approach through extensive experiments across various domains and demonstrate significant improvements over traditional methods in automating data alignment and optimizing transfer learning outcomes.