ANTS2 toolkit is being developed at LIP-Coimbra for simulation, optimization and processing of experimental data targeting mainly position-sensitive detectors with optical readout, such as gamma cameras, two phase noble gas detectors for searches of dark matter, and MSPC-based detectors for thermal neutrons. The main focus of the toolkit is position and energy reconstruction with access to a broad range of techniques including statistical methods and techniques based on machine learning (e.g. kNN searches and neural networks). A combination of a powerful custom B-spline library for detector response parameterization and fast reconstruction utilizing computations on Graphics Processing Unit makes ANTS2 an effective tool in real-time data processing. The simulation module of ANTS2 is based on CERN ROOT 3D navigation library (TGeo). A significant speedup of simulations was obtained by taking full advantage of the multithreading capabilities of TGeo and implementing custom tools for distribution of simulations over a grid of computers. The toolkit offers an interactive graphical as well as scripting (JavaScript and Python) interfaces. Based on the fast simulation module and the scripting interface, ANTS2 also give access to a set of tools for semi-supervised (highly automated) fast multi-parameter detector optimization.
Filipe Veloso e Pedro Costa