Search

Visual food recognition is emerging as an important application in dietary monitoring and management in recent years. Existing works use large backbone networks to achieve good performance. However, these networks are not able to be deployed on personal portable devices due to large size and computation cost. Some compact networks have been developed, however, their performance are usually lower than the large backbone networks. In view of this, this paper proposes a joint distillation framework that targets to achieve a high visual food recognition accuracy using a compact network. As opposed to the more traditional one-directional knowledge distillation methods, the proposed knowledge distillation framework trains both the large teacher network and the compact student network simultaneously. The framework introduces a new Multi-Layer Distillation (MLD) for simultaneous teacher-student learning at multiple layers of different abstraction. A novel Instance Activation Mapping (IAM) is proposed to jointly train the teacher and student networks using generated instance-level activation map that incorporates label information for each training image. Experimental results on the two benchmark datasets UECFood-256 and Food-101 show that the trained compact student network achieves state-of-the-art performance at 83.5% and 90.4%, respectively, while achieving more than 4 times deduction regarding network model size.

In recent years, palmprint recognition has made great progress and many methods have been put forward. The extraction of robust orientation features and finding efficient matching strategies are two key points for palmprint recognition. Traditional coding methods usually only use a dominant filter response to extract orientation features of palmprint images while not taking into account the other useful filter responses. Without increasing the number of filers, this study presents a modified Competitive Code to extract orientation features more accurately, which makes use of the relation between the filter responses. Besides, a distinctive extended eight-pixel neighbourhood method is proposed to select the sample points for matching by extracting the local features. At the matching stage, an effective fusion matching scheme with a double-layer image pyramid is designed to calculate the similarity between two palmprint images. Extensive experiments on four types of public palmprint databases show that the proposed method has excellent performance compared with the other state-of-the-art algorithms.

Ethnicity information is an integral part of human identity, and a useful identifier for various applications ranging from video surveillance, targeted advertisement to social media profiling. In recent years, Convolutional Neural Networks (CNNs) have shown state-of-the-art performance in many visual recognition problems. Currently, there are a few CNN-based approaches on ethnicity classification [1], [2]. However, the approaches suffer from the following limitations: (i) most face datasets do not include ethnicity information, and those with ethnicity information are typically small to medium in size, thereby they do not provide sufficient samples for training of CNNs from the scratch, and (ii) the CNN methods often treat ethnicity classification as a multi-class classification where the likelihood of each class label is generated. However, it does not utilize the intermediate activation functions of CNNs which provide rich hierarchical features to assist in ethnicity classification. In view of this, this paper proposes a new hybrid supervised learning method to perform ethnicity classification that uses both the strength of CNN as well as the rich features obtained from the network. The method combines the soft likelihood of CNN classification output with an image ranking engine that leverages on matching of the hierarchical features between the query and dataset images. A supervised Support Vector Machine (SVM) hybrid learning is developed to train the combined feature vectors to perform ethnicity classification. The performance of the proposed method is evaluated using a dataset consisting of Bangladeshi, Chinese and Indian ethnicity groups, and it outperforms the state-of-the-art methods [2], [3] by up to 3% in recognition accuracy.

Palmprint deformation, including rotation, displacement, swelling and so on, is a crucial factor for the palmprint recognition performance. The traditional algorithms mainly focus on the feature-based matching and the block-based matching. In this paper, a novel block-based matching palmprint recognition scheme is proposed. The sampling points on the template image are built by the harris corner detection algorithm. Furthermore, the corresponding sampling points on the query palmprint image are estimated by image pyramids algorithm for reducing the impact of distortion. At last, a matching method based on statistical offset is proposed to calculate the final matching score. Extensive experiments on public palmprint database shows that our method can achieve excellent recognition performance compared with many other classic methods.