BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Australia/Brisbane
X-LIC-LOCATION:Australia/Brisbane
BEGIN:STANDARD
DTSTART:19920301T030000
TZOFFSETFROM:+1100
TZOFFSETTO:+1000
TZNAME:AEST
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20191203T080501Z
LOCATION:Plaza Meeting Room P2
DTSTART;TZID=Australia/Brisbane:20191119T121800
DTEND;TZID=Australia/Brisbane:20191119T124500
UID:siggraphasia_SIGGRAPH Asia 2019_sess112_papers_318@linklings.com
SUMMARY:Automatically Translating Image Processing Libraries to Halide
DESCRIPTION:Technical Papers\n\nAutomatically Translating Image Processing
  Libraries to Halide\n\nAhmad, Ragan-Kelley, Cheung, Kamil\n\nThis paper p
 resents Dexter, a new tool that automatically translates image processing 
 functions from a low-level general-purpose language to a high-level domain
 -specific language (DSL), allowing them to leverage cross-platform optimiz
 ations enabled by DSLs. Rather than building a classical syntax-driven com
 piler to do this translation, Dexter leverages recent advances in program 
 synthesis and program verification, along with a new domain-specific synth
 esis algorithm, to translate C++ image processing code to the Halide DSL, 
 while guaranteeing semantic equivalence. This new synthesis algorithm scal
 es and generalizes to much larger and more complex functions than prior wo
 rk, including the ability to handle tiling, conditionals, and multi-stage 
 pipelines in the original low-level code. To demonstrate the effectiveness
  of our approach, we evaluate Dexter using real-world image processing fun
 ctions from Adobe Photoshop, a widely-used multi-platform image processing
  program. Our results show that Dexter can translate 264 out of 353 functi
 ons in our test set, with the original implementations ranging from 20 to 
 150 lines of code. By leveraging Halide’s advanced auto-scheduling capabil
 ities, we demonstrate median speedups of 7.03x and 4.52x for Dexter-transl
 ated functions as compared to the original implementations on Intel and AR
 M architectures, respectively.\n\nRegistration Category: Platinum Pass, Fu
 ll Conference Pass, Full Conference One-Day Pass
URL:https://sa2019.conference-program.com/presentation?id=papers_318&sess=
 sess112
END:VEVENT
END:VCALENDAR