Hybrid-Attention Based Decoupled Metric Learning for Zero-Shot Image Retrieval

Abstract

In zero-shot image retrieval (ZSIR) task, embedding learning becomes more attractive, however, many methods follow the traditional metric learning idea and omit the problems behind zero-shot settings. In this paper, we first emphasize the importance of learning visual discriminative metric and preventing the partial/selective learning behavior of learner in ZSIR, and then propose the Decoupled Metric Learning (DeML) framework to achieve these individually. Instead of coarsely optimizing an unified metric, we decouple it into multiple attention-specific parts so as to recurrently induce the discrimination and explicitly enhance the generalization. And they are mainly achieved by our object-attention module based on random walk graph propagation and the channel-attention module based on the adversary constraint, respectively. We demonstrate the necessity of addressing the vital problems in ZSIR on the popular benchmarks, outperforming the state-of-theart methods by a significant margin. Code is available at http://www.bhchen.cn * Corresponding author Figure 1: Differences between (a) unified metric learning and (b) DeML. Our DeML decouples the unified representations into multiple attention-specific learners so as to encourage the discrimination and generalization of the holistic metric.

Machine Learning-Based Highway Truck Commodity Classification Using Logo Data

“…To address this challenge, we adopt deep metric learning, which essentially 'learns to compare' image pairs by shrinking intra-class distance while expanding inter-class distance. Several deep metric learning approaches have been proposed and been applied in various applications, including face identification [18], person re-identification [19,20], and image retrieval [21]. Metric learning is known to have a stronger discriminative power since the inter-class distance is directly maximized during training.…”

A Single Historical Painting Super-Resolution via a Reference-Based Zero-Shot Network

“…The application of zero-shot in super-resolution, i.e, the ZSSR prescription [23], is among the most widely used models in superresolution and has gained increasing interest recently. In addition, the majority of zero-shot methods that have provided a large number of excellent results in recent years are mainly based on segmentation [39], emotion recognition [40], object detection [41], image retrieval [42][43][44][45], image classification [46][47][48] and intelligent learning in machines or robots [49]. In the ZSSR formalism, LR images are downsampled to generate many lower-resolution images (I = I 0 , I 1 , I 2 , ..., I n ), which serve as the HR supervision information called "HR fathers, " then, each HR father is downscaled by the required scale factor s to obtain the corresponding "LR sons.…”

Learning object-centric complementary features for zero-shot learning

“…Attention mechanism, widely studied and utilized in recent years, has been successfully applied in many computer vision tasks, such as image classification [6,12], semantic segmentation [27], and visual reasoning [28]. Inspired by the successful application of attention mechanisms in computer vision tasks, some works [30,31,32,40] attempt to introduce attention mechanisms to improve the discrimination of visual features in the few-shot and zero-shot learning task. A multi-attention network is proposed in [32] to solve one-shot learning problem.…”

“…A stacked semantic-guided attention model is proposed in [30] for fine-grained ZSL, which is capable of discovering discriminative subtle visual differences with the guidance of semantic. In the zero-shot retrieval task, [31] proposed a decoupled metric learning framework to learn discriminative visual attributes.…”