BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:DORAEMON T1-WC
DTSTART:20260903T010000Z
DTEND:20260903T020000Z
DTSTAMP:20261016T072400Z
UID:indico-event-10743@indico.slac.stanford.edu
DESCRIPTION:https://u-tokyo-ac-jp.zoom.us/j/98126050083\nLink to recording
 \n \nQuick recap\nThis meeting focused on planning the DORAEMON Thrust 1 
 (T1) water Cherenkov data challenge. Patrick explained the overall structu
 re of DORAEMON\, which aims to standardize data sets and create open data 
 challenges using AI models to improve physics analysis. The group discusse
 d Cesar's work using LUCiD\, a new differentiable simulator\, to generate 
 realistic water Cherenkov events that could be used for public data challe
 nges. Key decisions included whether to include detailed particle track se
 gmentation (case C) versus simpler event classification (case A)\, and the
  need to expand the current particle list to include gamma rays\, more pro
 ton decay modes\, and additional hadronic interactions. Benda suggested ad
 ding calibration and detector shift challenges to address data-MC shift is
 sues\, though implementation of complex detector variations like coil fail
 ures would require significant additional work. The team agreed to documen
 t their requirements in the shared Google Doc and meet again the following
  week to finalize proposals before the upcoming workshop.\nNext steps\nBen
 da Xu\n\nAdd ideas and proposals for calibration and detector parameter va
 riations (e.g.\, detector shifts\, data-MC consistency) to the Google Doc\
 , formalizing the discussion into a proposal for the DORAEMON data challen
 ge.\n\nPatrick\n\nAsk the DORAEMON group if there is an existing proposal 
 format/example to follow for the data challenge.\n\nTakuya\n\nPrepare a li
 st of particle data sets (including additional particles and event types n
 eeded) in a Google Doc or spreadsheet\, to be used as the final list for t
 he data challenge.\n\nCollaboration\n\nAll (Patrick\, Takuya\, Benda Xu\, 
 Shuoyu): Contribute to the Google Doc to define tasks\, metrics\, and data
 set definitions for the T1 Water Cherenkov data challenge proposal.\nAll: 
 Meet weekly at the same time next week to continue planning for the worksh
 op.\n\nSummary\nDORAEMON Project Planning Discussion\nPatrick and Takuya d
 iscussed the DORAEMON structure\, which has been divided into thrusts with
  water Cherenkov (WC) that still needs to be defined. Patrick explained th
 at much of the work has been happening on the US side with Cesar\, Omar\, 
 Junjie\, and Kazu\, but the team needs to determine what would be most use
 ful for Hyper-K.\nDORAEMON Physics AI Project\nPatrick explained the DORAE
 MON project\, which is based on an open data challenge\, aimed at organizi
 ng community efforts to solve physics problems using AI models. The projec
 t has three thrusts\, with Thrust 1 focusing on data-driven reconstruction
 \, which has seen the most development so far within Hyper-K. Patrick note
 d that while they previously attempted to standardize data sets by consoli
 dating them on a common cluster\, this hasn't been consistently adopted by
  the Hyper-K Reconstruction Group\, making it challenging to expand to the
  public community.\nHyper-K Physics Metrics Discussion\nPatrick discussed 
 a working document prepared by Cesar\, Omar\, and Kazu\, along with Benda 
 and himself\, outlining tasks and metrics for Hyper-K physics\, including 
 vertex and momentum reconstruction\, classification\, and interaction reco
 nstruction. He explained the importance of defining metrics for evaluating
  performance on these tasks\, particularly for multi-particle events\, usi
 ng parameters like IoU for pixel grouping and position\, direction\, and m
 omentum reconstruction errors. Patrick also reviewed Cesar's presentation 
 at NPML\, which demonstrated the use of the LUCiD differentiable simulator
  to generate realistic water Cherenkov events\, though the physics represe
 ntation needs further validation. The discussion highlighted the need to d
 efine detailed labels and segmentation approaches for multi-particle event
 s to improve physics analysis sensitivity.\nParticle Track Reconstruction 
 Approaches\nPatrick and Takuya discussed different approaches to particle 
 track reconstruction in Hyper-K data\, with Patrick noting that Mathias's 
 previous multi-ring segmentation work only considered whole tracks rather 
 than individual particle scattering and track segmentation. Takuya suggest
 ed that including more detailed track information (case C in Cesar's slide
 s) could help constrain uncertainties in high-energy analysis\, particular
 ly for pion scattering. The team agreed they need to make decisions about 
 which data generation and truth label approach to use and communicate this
  to the rest of the group for the data challenge.\nHyper-K Pilot Productio
 n Discussion\nThe team discussed a pilot production using super-K-like geo
 metry and considered switching to HK far detector and IWCD\, while also ma
 intaining interest in Super-K and WCTE existing experiments. They reviewed
  a list of simulated event types including single particles\, multi-partic
 le combinations\, and specialized samples\, questioning if the current lis
 t was complete and considering limitations in data size and organizational
  work for future challenges. The discussion highlighted missing gamma even
 ts and differences in multi-particle event creation method used by Mathias
  in HK. Takuya asked about the absence of muon+pizero events in the initia
 l set\, which is just because it was not a completely informed set yet.\nL
 UCiD Data Generation Strategy\nBenda Xu proposed using LUCiD to generate d
 ata for a MC consistency challenge by leveraging different phases of SK (S
 uper-Kamiokande) data\, including parameters from SK6 and SK7.5\, with the
  possibility of deliberately modifying the LUCiD setup to create more chal
 lenging scenarios. Patrick raised concerns about implementing missing phys
 ics in LUCiD\, particularly coil failures\, and questioned the feasibility
  of using SK data to improve LUCiD models while maintaining public access 
 to the generated datasets.\nDetector Model Development Discussion\nPatrick
  and Benda Xu discussed the development of a model to describe changes in 
 PMT response based on position in the detector and magnetic field\, which 
 they plan to implement eventually in general. They noted that while partic
 le and photon propagation are already implemented in their LUCiD model\, d
 etector variations and calibration shifts still need to be developed. The 
 team agreed to document all desired features in their Google Doc for the u
 pcoming workshop\, focusing on "blue skies" ideas without specific time/pe
 rsonpower constraints\, before prioritizing.\nWorkshop Task Definition Pla
 nning\nThe team discussed plans for defining tasks and datasets for an upc
 oming workshop and challenge proposal. Benda proposed using different dete
 ctor physics models (WCSim vs LUCiD) to create MC challenges\, while Takuy
 a will handle particle dataset definitions. Patrick emphasized the need to
  formalize their discussions into a proposal for DORAEMON\, including task
 s\, metrics\, and dataset definitions. The group agreed to meet weekly\, w
 ith Benda and Takuya confirming availability for next week's meeting.\n\nh
 ttps://indico.slac.stanford.edu/event/10743/
URL:https://indico.slac.stanford.edu/event/10743/
END:VEVENT
END:VCALENDAR