[{"data":1,"prerenderedAt":365},["ShallowReactive",2],{"project-9828962":3},{"id":4,"name":5,"line":6,"line_source":7,"title":8,"url":9,"domain":10,"maker":11,"posted_at":12,"year":13,"points":14,"comments":15,"rank_in_year":16,"shown_in_year":17,"top":18,"tier":19,"category":20,"subcategory":21,"format":22,"keywords":23,"reception":28,"sentiment":29,"stance_mix":30,"scored_comments":35,"aspects":36,"strip":44,"post_text":45,"what":46,"said":50,"maker_in_thread":130,"shape":141,"grid":145,"attempts":327,"same_week":328,"maker_shows":39,"maker_hidden":357,"reader_version":358,"card":359},"9828962","Trigrad","A novel image compression with interesting results","title","Show HN: Trigrad, a novel image compression with interesting results","http:\u002F\u002Fruarai.github.io\u002FTrigrad\u002F","ruarai.github.io","ruarai",1435978691,2015,175,67,104,2565,5,"notable","languages","image compression","desktop_app",[21,24,25,26,27],"triangulation","delaunay","lossy compression","gouraud shading","curious",-0.042,{"positive":31,"neutral":32,"mixed":33,"critical":34},6,26,0,8,40,{"idea":37,"docs":40,"performance":41,"execution":42},{"positive":38,"critical":39},2,1,{"positive":33,"critical":33},{"positive":33,"critical":18},{"positive":43,"critical":38},3,[18,33,39,33,33,39,39,43,39,18,18,18,33,43,43,43,33,33,33,43,33,33,43,39,43,43,39,43,33,39,33,43,43,33,39,43,18,43,43,18],null,{"problem":45,"audience":45,"pricing":45,"motivation":45,"time_to_build":45,"stack":47,"openness":48,"stage":49,"team_size":48},[],"unknown","prototype",{"praise":51,"objections":67,"requests":87,"questions":106,"alternatives":114},[52],{"value":53,"source":54,"evidence":55},"Interesting, working image compression idea that gradients handle well","commenter",[56,60,63],{"id":57,"by":58,"text":59},"9829601","kobigurk","This is the difference between someone who actually does something and academic work that claims to achieve something. This is half-done, but it WORKS and you can use it and…",{"id":61,"by":11,"text":62},"9829351","It's not very comparable to PNG, since they're designed for different types of imagery. I know currently Trigrad cannot handle text at all. In regards to your other comment, it's…",{"id":64,"by":65,"text":66},"9829442","ezdiy","Love how the algorithm is simple compared to iDCT based ones, very good job! > No text at all Or indeed anything which poorly maps to gradients. For this I'm thinking about…",[68,76],{"value":69,"source":54,"evidence":70},"Compares poorly to JPEG and PNG on photos and text",[71,75],{"id":72,"by":73,"text":74},"9829324","darkmighty","PNG is lossless, so I don't get the quality comparison. I don't think this approach can compete with JPEG and newer transform based variants for natural photos (event at edge…",{"id":61,"by":11,"text":62},{"value":77,"source":54,"evidence":78},"No comparison against existing state-of-the-art methods",[79,83],{"id":80,"by":81,"text":82},"9830508","cafebeen","I agree that it's nice to have this kind of write up and code, but \"works\" is a low bar to pass, especially in image compression. If this were written up as an academic paper, a…",{"id":84,"by":85,"text":86},"9829882","bastijn","I mean the author no harm, nor want to talk bad about his work. His work is very cool and I like the amount of information that he gives. Yet, this is far from comparable to…",[88,99],{"value":89,"source":54,"evidence":90},"Compare results against the Lenna benchmark and file sizes",[91,95],{"id":92,"by":93,"text":94},"9829607","rincewind","Can you try it with the lenna benchmark image and post the results please?",{"id":96,"by":97,"text":98},"9829992","userbinator","Nice pictures, but they should really have indicated the filesize for the 3000-sample version and given more details about this part: the samples can be saved and zipped up…",{"value":100,"source":54,"evidence":101},"Smoother interpolation or data-dependent triangulation",[102],{"id":103,"by":104,"text":105},"9832275","CyberDildonics","This is neat and the illustrations are great. A few things that will probably give large gains while being low hanging fruit: 1. There are triangle interpolation schemes out…",[107],{"value":108,"source":54,"evidence":109},"How does speed compare to other compression algorithms?",[110],{"id":111,"by":112,"text":113},"9829266","bdamos","How does the speed compare to other compression algorithms?",[115,123],{"value":116,"source":54,"evidence":117},"JPEG",[118,122],{"id":119,"by":120,"text":121},"9830049","mainguy","Not sure exactly, but wouldn't these images theoretically scale up better than jpeg? i.e. Making a 600x800 image out of a moderately compressed 300x400 seems like these would…",{"id":72,"by":73,"text":74},{"value":124,"source":54,"evidence":125},"Data-dependent triangulation papers by Lehner et al.",[126],{"id":127,"by":128,"text":129},"9829360","a_e_k","Garland and Heckbert had a nice algorithm for this sort of thing in their 1995 paper, \"Fast Polygonal Approximation of Terrains and Height Fields.\" The paper is mainly devoted to…",{"comments":34,"answered":31,"key":131},[132,135,138],{"id":133,"text":134},"9829699","Not sure if that applies for my purposes, since I'm not actually using linear interpolation barycentric coordinates (I don't think that's possible). The barycentric coordinates…",{"id":136,"text":137},"9829501","I tried your struct-of-array idea, and that's produced an okay improvement ~1%. Sorting them seems tough as the index of each value must match for each channel, so any sorting…",{"id":139,"text":140},"9829718","Wow, that works pretty well. I was mistaken in thinking that either the Dictionary class or the process of sampling would sort them. Mind if I merge that? Or you could submit a…",{"commenters":142,"max_depth":31,"maker_replies":34,"first_comment_mins":143,"lifespan_hours":144},39,116,91,[146,148,150,153,157,161,163,165,169,173,176,178,180,184,185,188,192,195,198,200,202,204,207,210,214,215,218,222,224,226,228,230,232,234,236,238,240,242,245,247,251,255,257,259,261,264,269,272,274,278,280,284,288,290,292,294,296,299,300,302,304,305,308,312,316,318,323],[57,33,18,147,58,33],"project\u002Fpositive\u002Fidea",[84,33,33,149,85,33],"commenter\u002Fneutral\u002Fnone",[151,33,33,149,152,33],"9830618","tjradcliffe",[154,33,39,155,156,33],"9831693","project\u002Fcritical\u002Fidea","pezzep",[158,33,43,159,160,33],"9832382","project\u002Fneutral\u002Fdocs","mc808",[162,33,33,149,85,33],"9848404",[164,33,33,149,97,33],"9829978",[166,33,33,167,168,33],"9830262","other\u002Fneutral\u002Fnone","rer0tsaz",[170,33,43,171,172,33],"9830929","project\u002Fneutral\u002Fperformance","nitrogen",[174,33,39,175,168,33],"9831042","project\u002Fcritical\u002Fperformance",[80,33,39,177,81,33],"project\u002Fcritical\u002Fexecution",[127,39,43,179,128,33],"project\u002Fneutral\u002Fidea",[181,39,43,182,183,33],"9829573","project\u002Fneutral\u002Fexecution","jacobolus",[133,39,33,182,11,39],[186,39,39,187,183,33],"9832071","maker\u002Fcritical\u002Fexecution",[189,38,18,190,191,33],"9830454","project\u002Fpositive\u002Fexecution","TheLoneWolfling",[193,38,18,190,194,33],"9830700","scottfr",[196,43,43,182,197,33],"9829453","pjtr",[136,43,33,199,11,39],"maker\u002Fneutral\u002Fexecution",[201,43,18,190,197,33],"9829632",[139,43,33,203,11,39],"maker\u002Fpositive\u002Fexecution",[205,43,43,206,197,33],"9829797","maker\u002Fneutral\u002Fnone",[208,43,43,206,209,33],"9830332","willvarfar",[211,212,33,167,213,33],"9830181",4,"arketyp",[96,18,43,159,97,33],[216,31,43,179,217,33],"9829770","JoshTriplett",[219,31,33,220,221,33],"9829812","commenter\u002Fneutral\u002Fexecution","1wd",[223,31,33,220,217,33],"9829926",[225,31,33,149,104,33],"9832256",[227,31,33,149,221,33],"9832547",[229,31,33,149,104,33],"9845830",[103,231,43,182,104,33],7,[233,231,43,182,191,33],"9832319",[235,231,33,167,128,33],"9832335",[237,231,33,167,191,33],"9832406",[239,231,33,167,128,33],"9832848",[241,231,33,167,191,33],"9836203",[243,34,43,179,244,33],"9830173","atinoda-kestrel",[246,34,43,179,128,33],"9832862",[248,249,39,175,250,33],"9829485",9,"slavik81",[252,253,43,182,254,33],"9829785",10,"JosephRedfern",[256,253,33,199,11,39],"9829809",[258,253,43,199,254,33],"9829853",[119,260,43,179,120,33],11,[262,260,39,175,263,33],"9830681","evan_",[265,266,43,267,268,33],"9829495",12,"maker\u002Fneutral\u002Fdocs","joshu",[270,266,33,271,221,33],"9829542","other\u002Fneutral\u002Fdocs",[273,266,33,149,268,33],"9843897",[275,276,39,175,277,33],"9830213",13,"tgb",[279,276,43,182,191,33],"9830442",[281,282,33,167,283,33],"9829639",14,"return0",[285,286,43,182,287,33],"9829741",15,"coolvision",[92,289,43,206,93,33],16,[291,289,33,175,11,39],"9829709",[111,293,43,171,112,33],17,[295,293,33,171,11,39],"9829282",[297,293,43,171,298,33],"9829312","ekianjo",[72,293,39,175,73,33],[301,293,43,171,298,33],"9829342",[61,293,33,303,11,39],"maker\u002Fmixed\u002Fperformance",[64,293,18,147,65,33],[306,293,43,171,307,33],"9829336","birdsbolt",[309,310,33,167,311,33],"9830115",18,"imaginenore",[313,314,43,182,315,33],"9829502",19,"boterock",[317,314,33,182,11,39],"9829767",[319,320,18,321,322,33],"9829304",20,"project\u002Fpositive\u002Fnone","akhilcacharya",[324,325,43,171,326,33],"9829768",21,"octatoan",[],[329,334,339,344,348,352],{"id":330,"name":331,"points":332,"category":333},"9838955","“Who is hiring",546,"productivity",{"id":335,"name":336,"points":337,"category":338},"9839712","YouTube Filesystem",185,"dev_tools",{"id":340,"name":341,"points":342,"category":343},"9830531","Silicon Valley Dictionary",158,"learning_science",{"id":345,"name":346,"points":347,"category":20},"9845763","node-rdpjs",140,{"id":349,"name":350,"points":351,"category":338},"9843268","ttystudio",118,{"id":353,"name":354,"points":355,"category":356},"9820521","Ironclad",105,"web_saas",false,"v2",{"id":360,"name":5,"line":6,"ai":357,"cat":20,"year":13,"points":14,"comments":15,"rec":28,"s":361,"top":18,"kw":362,"maker":11,"domain":10,"tier":19,"attempts":33,"strip":363,"url":9,"mix":364,"praise":53,"objection":69,"sub":21},9828962,-0.04,[21,24,25,26,27],"5010011315550333000300313313010330135335",[31,32,33,34],1793990927137]