Search

Most researchers focus on designing accurate crowd counting models with heavy parameters and computations but ignore the resource burden during the model deployment. A real-world scenario demands an efficient counting model with low-latency and high-performance. Knowledge distillation provides an elegant way to transfer knowledge from a complicated teacher model to a compact student model while maintaining accuracy. However, the student model receives the wrong guidance with the supervision of the teacher model due to the inaccurate information understood by the teacher in some cases. In this paper, we propose a dual-knowledge distillation (DKD) framework, which aims to reduce the side effects of the teacher model and transfer hierarchical knowledge to obtain a more efficient counting model. First, the student model is initialized with global information transferred by the teacher model via adaptive perspectives. Then, the self-knowledge distillation forces the student model to learn the knowledge by itself, based on intermediate feature maps and target map. Specifically, the optimal transport distance is utilized to measure the difference of feature maps between the teacher and the student to perform the distribution alignment of the counting area. Extensive experiments are conducted on four challenging datasets, demonstrating the superiority of DKD. When there are only approximately 6% of the parameters and computations from the original models, the student model achieves a faster and more accurate counting performance as the teacher model even surpasses it. [ABSTRACT FROM AUTHOR]

Unsupervised domain adaptation (UDA) methods have become a research hotspot in remote sensing scene classification to reduce dependence on labeled samples. However, most current methods focus on extracting domain invariant features, ignoring the problem of large intraclass differences and the imbalanced sample numbers between categories in remote sensing images. To address these issues, we propose a remote sensing scene domain adaptive method based on middle-layer feature extraction and nuclear norm maximization (MFE-NM). In the MFE module, the middle-layer features of the feature extractor are randomly extracted and processed. Since the receptive field of the middle-layer features is smaller and the resolution is higher, the effective use of the middle-layer features can reduce the impact of image feature heterogeneity caused by large intraclass differences in remote sensing images. In addition, it can be concluded that the constrained nuclear norm can simultaneously improve the prediction diversity and discriminability of the model through theoretical derivation. Therefore, the NM module is proposed to solve the problem of reduced prediction diversity caused by entropy minimization methods when dealing with scene classification problems with imbalanced sample numbers between categories. Extensive experiments and analyses on three public remote sensing datasets demonstrate the effectiveness and competitiveness of our proposed method. [ABSTRACT FROM AUTHOR]

Contrastive learning has revolutionized the field of computer vision, learning rich representations from unlabeled data, which generalize well to diverse vision tasks. Consequently, it has become increasingly important to explain these approaches and understand their inner workings mechanisms. Given that contrastive models are trained with interdependent and interacting inputs and aim to learn invariance through data augmentation, the existing methods for explaining single-image systems (e.g., image classification models) are inadequate as they fail to account for these factors and typically assume independent inputs. Additionally, there is a lack of evaluation metrics designed to assess pairs of explanations, and no analytical studies have been conducted to investigate the effectiveness of different techniques used to explaining contrastive learning. In this work, we design visual explanation methods that contribute towards understanding similarity learning tasks from pairs of images. We further adapt existing metrics, used to evaluate visual explanations of image classification systems, to suit pairs of explanations and evaluate our proposed methods with these metrics. Finally, we present a thorough analysis of visual explainability methods for contrastive learning, establish their correlation with downstream tasks and demonstrate the potential of our approaches to investigate their merits and drawbacks. [ABSTRACT FROM AUTHOR]