Generic human detectors perform poorly in application scenes in which conditions are significantly different from those of the benchmark data sets. Based on the assumption that only a limited number of labeled examples are available, we propose a variant semi-supervised boosting approach for improving scene adaptiveness by utilizing unlabeled data. Specifically, we train a max-margin-based model as an initial detector for new example collection, instead of using generic detectors, and then a better model is trained via boosting in which the newly obtained examples influence the training process through their similarities to the labeled examples. Since the widely used human descriptors are usually high dimensional and redundant, we employ a graph-based method to determine the weight representing the importance of each feature, such that the weighted similarity measurement leads to a performance gain. In the experiments, the effectiveness of the proposed approach “Variant SemiBoost” is demonstrated and state-of-the-art performance on challenging data sets is achieved.

The difference between sample distributions of public data sets and specific scenes can be very significant. As a result, the deployment of generic human detectors in real-world scenes most often leads to sub-optimal detection performance. To avoid the labor-intensive task of manual annotations, we propose a semi-supervised approach for training deep convolutional networks on partially labeled data. To exploit a large amount of unlabeled target data, the knowledge learnt from public data sets is transferred to new model training by adapting an auxiliary detector to the target scene. We hypothesize that the components of the auxiliary detector capture essential human characteristics useful for constructing a scene-adapted detector. A selective ensemble algorithm is proposed to select a subset of the components relevant to the target scene for recombination. The resulting model is applied for collecting high-confidence samples from unlabeled target data. Furthermore, a deep convolutional network is trained by progressively labeling and selecting new training samples in a self-paced way. The detailed experimental evaluation verifies the effectiveness and superiority of the proposed approach in scene-specific human detection.

In a semi-supervised setting, direct training of a deep discriminative model on partially labeled images often suffers from overfitting and poor performance, because only a small number of labeled images are available, and errors in label propagation are, in many cases, inevitable. In this paper, we introduce an auxiliary clustering task to explore the structure of the image data, and judiciously weigh unlabeled data to alleviate the influence of ambiguous data on model training. For this purpose, we propose a cross-task network composed of two streams to jointly learn two tasks: classification and clustering. Based on the model predictions, a large number of pairwise constraints can be generated from unlabeled images, and are fed to the clustering stream. Since pairwise constraints encode weak supervision information, the clustering is tolerant of errors in labeling. Unlabeled images are weighted according to the distances to the clusters discovered, and a better discriminative model is trained on the classification stream associated with a weighted softmax loss. Furthermore, a self-paced learning paradigm is adopted to gradually train our deep model from easy examples to difficult ones. Experimental results on widely used image classification datasets confirm the effectiveness and superiority of the proposed approach.

Agents are difficult to test because it is notoriously complicated to observe their proactive, autonomous and non-deterministic behaviours and hard to judge their correctness in dynamic environments. This paper proposes a specification-based test automation framework and presents a tool called CATest for testing multi-agent systems (MAS). The agent-based formal specification language SLABS plays three roles in the framework. First, it is used to guide the instrumentation of the agent under test so that its behaviour can be observed and recorded systematically. Second, the correctness of agent's behaviours recorded during test executions are automatically checked against the formal specifications. Finally, the test adequacy is measured by the coverage of the specification and determined according to a set of adequacy criteria specifically designed for testing MAS. An experiment with the tool has demonstrated its capability of detecting faults in MAS.

Emergent behavior is an essential feature in multi-agent systems and plays a significant role in the applications of agent technology. Because of the huge gap between individual agents' behaviors and those of the whole system, specifying and reasoning about emergent behaviors are notoriously difficult. Simulation has been the essential method to study emergent behaviors in multi-agent systems. In this paper, we report an experimental study of the emergent behaviors of self-organized agent communities, in which emergent behaviors play a crucial role. The experiments confirmed the results of a theoretical analysis of agent communities using a formal theory called scenario calculus. It further provided insight into the dynamic features of the system that were very difficult to obtain by using formal logic, such as the speed of convergence to the emergent states and the relationships between the convergence time and various parameters of self-organized agent communities systems.