Zuping Zhang

Mohammed Al-Habib, Zuping Zhang, Abdulrahman Noman

Vision Transformers (ViTs) have shown significant promise in computer vision applications. However, their performance in few-shot learning is limited by challenges in refining token-level interactions, struggling with limited training data, and developing a strong inductive bias. Existing methods often depend on inflexible token matching or basic similarity measures, which limit the effective incorporation of global context and localized feature refinement. To address these challenges, we propose Bi-Level Adaptive Token Refinement for Few-Shot Transformers (BATR-FST), a two-stage approach that progressively improves token representations and maintains a robust inductive bias for few-shot classification. During the pre-training phase, Masked Image Modeling (MIM) provides Vision Transformers (ViTs) with transferable patch-level representations by recreating masked image regions, providing a robust basis for subsequent adaptation. In the meta-fine-tuning phase, BATR-FST incorporates a Bi-Level Adaptive Token Refinement module that utilizes Token Clustering to capture localized interactions, Uncertainty-Aware Token Weighting to prioritize dependable features, and a Bi-Level Attention mechanism to balance intra-cluster and inter-cluster relationships, thereby facilitating thorough token refinement. Furthermore, Graph Token Propagation ensures semantic consistency between support and query instances, while a Class Separation Penalty preserves different class borders, enhancing discriminative capability. Extensive experiments on three benchmark few-shot datasets demonstrate that BATR-FST achieves superior results in both 1-shot and 5-shot scenarios and improves the few-shot classification via transformers.

Chengyuan Zhang, Yunwu Lin, Lei Zhu et al.

This paper aims to solve the problem of large-scale video retrieval by a query image. Firstly, we define the problem of top-$k$ image to video query. Then, we combine the merits of convolutional neural networks(CNN for short) and Bag of Visual Word(BoVW for short) module to design a model for video frames information extraction and representation. In order to meet the requirements of large-scale video retrieval, we proposed a visual weighted inverted index(VWII for short) and related algorithm to improve the efficiency and accuracy of retrieval process. Comprehensive experiments show that our proposed technique achieves substantial improvements (up to an order of magnitude speed up) over the state-of-the-art techniques with similar accuracy.

Chengyuan Zhang, Ruipeng Chen, Lei Zhu et al.

In the big data era, massive amount of multimedia data with geo-tags has been generated and collected by mobile smart devices equipped with mobile communications module and position sensor module. This trend has put forward higher request on large-scale of geo-multimedia data retrieval. Spatial similarity join is one of the important problem in the area of spatial database. Previous works focused on textual document with geo-tags, rather than geo-multimedia data such as geo-images. In this paper, we study a novel search problem named spatial visual similarity join (SVS-JOIN for short), which aims to find similar geo-image pairs in both the aspects of geo-location and visual content. We propose the definition of SVS-JOIN at the first time and present how to measure geographical similarity and visual similarity. Then we introduce a baseline inspired by the method for textual similarity join and a extension named SVS-JOIN$_G$ which applies spatial grid strategy to improve the efficiency. To further improve the performance of search, we develop a novel approach called SVS-JOIN$_Q$ which utilizes a quadtree and a global inverted index. Experimental evaluations on real geo-image datasets demonstrate that our solution has a really high performance.

Kamal Al-Sabahi, Zhang Zuping, Yang Kang

Sequence generative models with RNN variants, such as LSTM, GRU, show promising performance on abstractive document summarization. However, they still have some issues that limit their performance, especially while deal-ing with long sequences. One of the issues is that, to the best of our knowledge, all current models employ a unidirectional decoder, which reasons only about the past and still limited to retain future context while giving a prediction. This makes these models suffer on their own by generating unbalanced outputs. Moreover, unidirec-tional attention-based document summarization can only capture partial aspects of attentional regularities due to the inherited challenges in document summarization. To this end, we propose an end-to-end trainable bidirectional RNN model to tackle the aforementioned issues. The model has a bidirectional encoder-decoder architecture; in which the encoder and the decoder are bidirectional LSTMs. The forward decoder is initialized with the last hidden state of the backward encoder while the backward decoder is initialized with the last hidden state of the for-ward encoder. In addition, a bidirectional beam search mechanism is proposed as an approximate inference algo-rithm for generating the output summaries from the bidi-rectional model. This enables the model to reason about the past and future and to generate balanced outputs as a result. Experimental results on CNN / Daily Mail dataset show that the proposed model outperforms the current abstractive state-of-the-art models by a considerable mar-gin.

Chengyuan Zhang, Yunwu Lin, Lei Zhu et al.

Online social networking techniques and large-scale multimedia systems are developing rapidly, which not only has brought great convenience to our daily life, but generated, collected, and stored large-scale multimedia data. This trend has put forward higher requirements and greater challenges on massive multimedia data retrieval. In this paper, we investigate the problem of image similarity measurement which is used to lots of applications. At first we propose the definition of similarity measurement of images and the related notions. Based on it we present a novel basic method of similarity measurement named SMIN. To improve the performance of calculation, we propose a novel indexing structure called SMI Temp Index (SMII for short). Besides, we establish an index of potential similar visual words off-line to solve to problem that the index cannot be reused. Experimental evaluations on two real image datasets demonstrate that our solution outperforms state-of-the-art method.

Chengyuan Zhang, Yunwu Lin, Lei Zhu et al.

With the proliferation of online social networking services and mobile smart devices equipped with mobile communications module and position sensor module, massive amount of multimedia data has been collected, stored and shared. This trend has put forward higher request on massive multimedia data retrieval. In this paper, we investigate a novel spatial query named region of visual interests query (RoVIQ), which aims to search users containing geographical information and visual words. Three baseline methods are presented to introduce how to exploit existing techniques to address this problem. Then we propose the definition of this query and related notions at the first time. To improve the performance of query, we propose a novel spatial indexing structure called quadtree based inverted visual index which is a combination of quadtree, inverted index and visual words. Based on it, we design a efficient search algorithm named region of visual interests search to support RoVIQ. Experimental evaluations on real geo-image datasets demonstrate that our solution outperforms state-of-the-art method.

Chengyuan Zhang, Kesheng Cheng, Lei Zhu et al.

With the rapid development of mobile Internet and cloud computing technology, large-scale multimedia data, e.g., texts, images, audio and videos have been generated, collected, stored and shared. In this paper, we propose a novel query problem named continuous top-$k$ geo-image query on road network which aims to search out a set of geo-visual objects based on road network distance proximity and visual content similarity. Existing approaches for spatial textual query and geo-image query cannot address this problem effectively because they do not consider both of visual content similarity and road network distance proximity on road network. In order to address this challenge effectively and efficiently, firstly we propose the definition of geo-visual objects and continuous top-$k$ geo-visual objects query on road network, then develop a score function for search. To improve the query efficiency in a large-scale road network, we propose the search algorithm named geo-visual search on road network based on a novel hybrid indexing framework called VIG-Tree, which combines G-Tree and visual inverted index technique. In addition, an important notion named safe interval and results updating rule are proposed, and based on them we develop an efficient algorithm named moving monitor algorithm to solve continuous query. Experimental evaluation on real multimedia dataset and road network dataset illustrates that our solution outperforms state-of-the-art method.

Kamal Al-Sabahi, Zhang Zuping, Mohammed Nadher

The recent advance in neural network architecture and training algorithms have shown the effectiveness of representation learning. The neural network-based models generate better representation than the traditional ones. They have the ability to automatically learn the distributed representation for sentences and documents. To this end, we proposed a novel model that addresses several issues that are not adequately modeled by the previously proposed models, such as the memory problem and incorporating the knowledge of document structure. Our model uses a hierarchical structured self-attention mechanism to create the sentence and document embeddings. This architecture mirrors the hierarchical structure of the document and in turn enables us to obtain better feature representation. The attention mechanism provides extra source of information to guide the summary extraction. The new model treated the summarization task as a classification problem in which the model computes the respective probabilities of sentence-summary membership. The model predictions are broken up by several features such as information content, salience, novelty and positional representation. The proposed model was evaluated on two well-known datasets, the CNN / Daily Mail, and DUC 2002. The experimental results show that our model outperforms the current extractive state-of-the-art by a considerable margin.