RESUMO
To the problem of the complex pre-processing and post-processing to obtain head-position existing in the current crowd localization method using pseudo boundary box and pre-designed positioning map, this work proposes an end-to-end crowd localization framework named WSITrans, which reformulates the weakly-supervised crowd localization problem based on Transformer and implements crowd counting. Specifically, we first perform global maximum pooling (GMP) after each stage of pure Transformer, which can extract and retain more detail of heads. In addition, we design a binarization module that binarizes the output features of the decoder and fuses the confidence score to obtain more accurate confidence score. Finally, extensive experiments demonstrate that the proposed method achieves significant improvement on three challenging benchmarks. It is worth mentioning that the WSITrans improves F1-measure by 4.0%.
Assuntos
Benchmarking , Fontes de Energia Elétrica , Processos MentaisRESUMO
Background: Spectrum-based Fault localization have proven to be useful in the process of software testing and debugging. However, how to improve the effectiveness of software fault localization has always been a research hot spot in the field of software engineering. Dynamic slicing can extract program dependencies under certain conditions. Thus, this technology is expected to benefit for locating fault. Methods: We propose an improved dynamic slicing for spectrum-based fault localization under a general framework. We first obtain the dynamic slice of program execution. Secondly, we construct a mixed slice spectrum matrix from the dynamic slice of each test case and the corresponding test results. Finally, we compute the suspiciousness value of each statement in the mixed-slice spectram matrix. Results: To verify the performance of our method, we conduct an empirical study on 15 widely used open-source programs. Experimental results show that our approach achieves significant improvement than the compared techniques. Conclusions: Our approach can reduce approximately 1% to 17.79% of the average cost of code examined significantly.