felicepantaleo/cms-data-RecoHGCal-TICL

Data for the CMSSW RecoHGCal/TICL package

Quicklinks

• TICL:
  • Reconstruction example
• TensorFlow:
  • C++ docs
  • CMSSW interface
  • PhysicsTools/TensorFlow

Models

• tf_models/energy_id_v*.pb: TensorFlow model for trackster energy regression and particle ID.
  • v0: Simple CNN-based approach. The neutral pion, neutral hadron, ambiguous and unknown probabilities are set to a constant value of 0. See the talk at the Reco/AT meeting for more info. Input and output tensors:
    • "input": Input tensor with dimension batch x 50 (layers) x 10 (clusters) x 3 (features).
    • "output/id_probabilities": Output tensor with dimension batch x 8 representing particle ID "probabilities" (from a softmax output). The probabiltities refer to photon, electron, muon, neutral pion, charged hadron, neutral hadron, ambiguous and unknown cases (in that order).
    • "output/regressed_energy": Output tensor with dimension batch x 1 representing the regressed energy value for the trackster.