24 December 2024 in Pattern Recognition Letters
22 March 2021 in Library Hi Tech
Purpose: The SARS-CoV-2 pandemic has caused a major impact on worldwide public health and economics. The lessons learned from the successful attempts to contain the pandemic escalation revealed that the wise usage of contact tracing and information systems can widely help the containment work of any contagious disease. In this context, this paper investigates other researches on this domain, as well as the main issues related to the practical implementation of such systems and specifies a technical solution.Design/methodology/approach: The proposed solution is based on the automatic identification of relevant contacts between infected or suspected people with susceptible people; inference of contamination risk based on symptoms history, user navigation records and contact information; real-time georeferenced information of population density of infected or suspect people; and automatic individual social distancing recommendation calculated through the individual contamination risk and the worsening of clinical condition risk.Findings: The solution was specified, prototyped and evaluated by potential users and health authorities. The proposed solution has the potential of becoming a reference on how to coordinate the efforts of health authorities and the population on epidemic control.Originality/value: This paper proposed an original information system for epidemic control which was applied for the SARS-CoV-2 pandemic and could be easily extended to other epidemics.
25 June 2026 in ACM Transactions on Architecture and Code Optimization
Deep Learning (DL) models are at the core of a growing number of applications, making fast, low-latency execution across diverse device architectures both a critical requirement and a challenge. DL compilers, such as TVM, address this challenge by automatically translating high-level models into optimized low-level code that effectively exploits device architectures. However, search-space-based algorithms face difficulties in exploring the vast optimization sequence space, often resulting in lengthy compilation times that significantly impact the design cycle. This article introduces the Task Graph Caching (TGC) algorithm, 1 which aims to reduce the high compilation time while preserving the quality of the code generated by TVM. In particular, TGC enhances TVM auto-tuning by exploiting the fact that similar DL subgraphs appear both within and across models, thus enabling optimization sequences discovered in past compilations to guide future executions. To achieve this, TGC introduces a cache structure that stores high-performance optimization sequences found in previous TVM executions. This information is then used to seed the population of TVM evolutionary search, avoiding redundant exploration of the optimization space and accelerating convergence. Experimental results on twelve DL models show that TGC can significantly speed up the search for efficient optimization sequences, reducing auto-tuning time by up to 2.89× for Ansor and 3.13× for MetaSchedule on CPU. Moreover, TGC reduces auto-tuning time by up to 3.25× for MetaSchedule on GPU. Furthermore, on average, TGC can maintain the inference time achieved by the default TVM, making it a promising solution for accelerating the compilation of DL models.