BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//pretalx.com//juliacon2024//speaker//XN93M7
BEGIN:VTIMEZONE
TZID:CET
BEGIN:STANDARD
DTSTART:20001029T040000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10
TZNAME:CET
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20000326T030000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3
TZNAME:CEST
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
UID:pretalx-juliacon2024-NVFAVL@pretalx.com
DTSTART;TZID=CET:20240710T105600
DTEND;TZID=CET:20240710T105900
DESCRIPTION:I spent about one year leveraging Julia for DL research in Comp
 uter Vision\, including using `Flux.jl`\, `FastAI.jl`\, `Metalhead.jl`\, a
 nd loading python models through `PyChainCall.jl`. Happy to discuss my exp
 erience.
DTSTAMP:20260310T064557Z
LOCATION:Method (1.5)
SUMMARY:My Experience With Deep Learning Research in Julia. - Romeo Valenti
 n
URL:https://pretalx.com/juliacon2024/talk/NVFAVL/
END:VEVENT
BEGIN:VEVENT
UID:pretalx-juliacon2024-V8VGK9@pretalx.com
DTSTART;TZID=CET:20240710T163000
DTEND;TZID=CET:20240710T164000
DESCRIPTION:We present `KSVD.jl`\, an extremely fast implementation of the 
 K-SVD algorithm\, including extensive single-core optimizations\, shared-s
 tate multithreading\, pipelined GPU offloading\, and an optional distribut
 ed executor. With this implementation\, we are able to outperform existing
  numpy-based implementations by ~100x and scale to datasets with millions 
 of samples.
DTSTAMP:20260310T064557Z
LOCATION:Function (4.1)
SUMMARY:KSVD.jl: A case study in performance optimization. - Romeo Valentin
URL:https://pretalx.com/juliacon2024/talk/V8VGK9/
END:VEVENT
BEGIN:VEVENT
UID:pretalx-juliacon2024-MVFVGB@pretalx.com
DTSTART;TZID=CET:20240710T171000
DTEND;TZID=CET:20240710T172000
DESCRIPTION:We present `RunwayPNPSolve.jl`\, a framework for uncertainty-aw
 are pose estimation for visual landing applications with multiple methods 
 including real-time least-squares minimization + resampling\, Monte-Carlo 
 Markov-Chain\, and a linear approximation\, by leveraging the existing Jul
 ia package ecosystem.\nThe package further provides a framework of useful 
 primitives to build simultaneously differentiable\, unitful and coordinate
 -system-aware interfaces\, and an interactive visualization pipeline.
DTSTAMP:20260310T064557Z
LOCATION:While Loop (4.2)
SUMMARY:RunwayPNPSolve.jl: Uncertainty-Aware Pose Estimation. - Romeo Valen
 tin
URL:https://pretalx.com/juliacon2024/talk/MVFVGB/
END:VEVENT
BEGIN:VEVENT
UID:pretalx-juliacon2024-VJEVSJ@pretalx.com
DTSTART;TZID=CET:20240710T190000
DTEND;TZID=CET:20240710T193000
DESCRIPTION:We present `ThreadedDenseSparseMul.jl`\, a library that efficie
 ntly computes dense-sparse multiplications and outperform competing packag
 es (and `Base.SparseArrays`) in about 20 lines of code (for the basic func
 tionality) by leveraging `Polyester.jl`. We further discuss the effect of 
 Julia's memory layout on the performance and analyze the influence of diff
 erent threading models.
DTSTAMP:20260310T064557Z
LOCATION:Method (1.5)
SUMMARY:ThreadedDenseSparseMul.jl: Multi-threaded Dense-Sparse Matmul. - Ro
 meo Valentin
URL:https://pretalx.com/juliacon2024/talk/VJEVSJ/
END:VEVENT
END:VCALENDAR