Image captioning is a fundamental task which requires semantic understanding of images and the ability of generating description sentences with proper and correct structure. In consideration of the problem that langua...
详细信息
ISBN:
(纸本)9781509060689
Image captioning is a fundamental task which requires semantic understanding of images and the ability of generating description sentences with proper and correct structure. In consideration of the problem that language models are always shallow in modern image caption frameworks, a deep residual recurrent neural network is proposed in this work with the following two contributions. First, an easy-to-train deep stacked Long Short Term Memory (LSTM) language model is designed to learn the residual function of output distributions by adding identity mappings to multi-layer LSTMs. Second, in order to overcome the over-fitting problem caused by larger-scale parameters in deeper LSTM networks, a novel temporal Dropout method is proposed into LSTM. The experimental results on the benchmark MSCOCO and Flickr30K datasets demonstrate that the proposed model achieves the state-of-the-art performances with 101.1 in CIDEr on MSCOCO and 22.9 in B-4 on Flickr30K, respectively.
Real-time wireless sensor networks in the industrial settings usually consist of tens of nodes. Seldom do we see a network of over 100 wireless nodes. The smaller size is usually sufficient as a typical plane unit is ...
In this paper, a large-scale human action recognition system is proposed which is built upon the combination of the rising big data processing technology Spark and the powerful Graphics Processing Unit (GPU) in order ...
详细信息
The password pattern has become a widely used mobile authentication method. However, there are still some potential security problems since the passwords are easy to be cracked by some malicious software. An important...
详细信息
A novel approach related to query expansion is proposed to improve image retrieval performance. The proposed approach investigates the problem that not all of the visual features extracted from images are appropriate ...
详细信息
In the classical computation theory, the language of a system features the computational behavior of the system but it does not distinguish the determinism and nondeterminism of actions. However, Milner found that the...
详细信息
Recently,finite-time and fixed-time stability has been a hot topic in the study of dynamical *** this paper,we will provide an insightful view on this issue and explore the intrinsic property by mathematical *** theor...
详细信息
ISBN:
(纸本)9781509009107
Recently,finite-time and fixed-time stability has been a hot topic in the study of dynamical *** this paper,we will provide an insightful view on this issue and explore the intrinsic property by mathematical *** theorems and corollaries show how to find the settling ***,we apply the corresponding theory on the finite-time and fixed-time synchronization problem of nonlinearly coupled systems and a numerical simulation is also given to demonstrate its efficiency.
Pythagorean fuzzy sets provide stronger ability than intuitionistic fuzzy sets to model uncertainty information and knowledge, but little effort has been paid to conflict analysis of Pythagorean fuzzy information syst...
详细信息
Inter-organizational workflow nets(IWF-nets) can well model the interactions among multiple processes by sending/receiving messages. Compatibility and weak compatibility are crucial properties for *** latter guarantee...
详细信息
Inter-organizational workflow nets(IWF-nets) can well model the interactions among multiple processes by sending/receiving messages. Compatibility and weak compatibility are crucial properties for *** latter guarantees that a system is deadlock-free and livelock-free while the former also guarantees that it has no dead tasks. Our previous work proved that the(weak) compatibility problem is PSPACE-complete for safe IWF-nets. This paper defines a class of IWF-nets in which some simple circuits are allowed. Necessary and sufficient conditions are presented to decide compatibility and weak compatibility for this class, and they are dependent on the net structures only. Algorithms are developed based on these conditions. In addition, we show that the traditional net structures like siphon cannot be easily used to decide the(weak) compatibility of IWF-nets.
With the development of the Internet, a lot of people trapped in the network, especially the adolescent depending on the network game and disturbing their normal life. 579 freshmen participated in this work who were c...
详细信息
ISBN:
(纸本)9781509040940
With the development of the Internet, a lot of people trapped in the network, especially the adolescent depending on the network game and disturbing their normal life. 579 freshmen participated in this work who were collected the personality questionnaires in the first week they came in university and their grades points average (GPA) after half year. The questionnaires were including Self-Control (SCS), Barratt impulse Inventory (BIS) and Chinese Big Five Personality (CBF). This work used multi-clustering algorithms to construct the models of predicting for Internet game disorder (IGD) risk, including FCM, K-means, and Hierarchical for training model. This is the first try to predict the risk of IGD by personality traits. The results shown the questionnaire data were well separated by different clustering algorithms into three groups who were shared the analogous personality traits which has a relationship with the behavior of IGD. But compared to the GPA of each group, the efficiency of the prediction model seems not so satisfactory. There need more efforts to optimized the model in the future.
暂无评论