[{"data":1,"prerenderedAt":470},["ShallowReactive",2],{"project-5045621":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":27,"sentiment":28,"stance_mix":29,"scored_comments":33,"aspects":34,"strip":46,"post_text":47,"what":48,"said":82,"maker_in_thread":210,"shape":220,"grid":224,"attempts":433,"same_week":434,"maker_shows":39,"maker_hidden":462,"reader_version":463,"card":464},"5045621","Dextro","API for detecting people, cars, and everyday objects in images","title","Show HN: API for detecting people, cars, and everyday objects in images","https:\u002F\u002Fwww.dextrorobotics.com","dextrorobotics.com","jluan",1357946414,2013,179,72,86,1739,5,"notable","ai","computer vision api","api",[24,25,21,26],"object detection","image recognition","face detection","roasted",-0.337,{"positive":30,"neutral":31,"mixed":30,"critical":32},6,14,27,53,{"idea":35,"performance":38,"execution":41,"docs":42,"pricing":44,"privacy":45},{"positive":36,"critical":37},3,0,{"positive":39,"critical":40},1,23,{"positive":37,"critical":39},{"positive":37,"critical":43},2,{"positive":39,"critical":37},{"positive":37,"critical":39},[18,36,39,18,39,37,39,39,37,37,37,39,36,36,39,39,39,36,39,36,39,39,36,39,39,39,39,36,39,36,39,36,39,36,36,39,36,36,39,36],null,{"problem":49,"audience":55,"pricing":67,"motivation":47,"time_to_build":47,"stack":79,"openness":80,"stage":81,"team_size":80},{"value":50,"source":51,"evidence":52},"Detecting people, cars and objects in images via an API","maker",[53],{"id":54},"post",{"value":56,"source":57,"evidence":58},"Developers building apps that need image detection","inferred",[59,63],{"id":60,"by":61,"text":62},"5046390","ank286","This is not Pedro Felzenszwalb's discriminative part-model algorithm. This is simple AdaBoost. The authors have labeled a bunch of datasets (1000s of them) and are able to detect…",{"id":64,"by":65,"text":66},"5046444","TommyDANGerous","That is amazingly awesome. Glad it can integrate with Ruby and Python. I haven't even read the whole info and I already signed up.",{"value":68,"source":69,"evidence":70},"Small free usage tier, paid usage beyond it","commenter",[71,75],{"id":72,"by":73,"text":74},"5045692","IceyEC","That's great! I like that you guys are offering a small free usage tier!",{"id":76,"by":77,"text":78},"5045776","steeve","As a long time CV enthusiast, I applaude the tech and the way you guys make it \"just work\". However for any serious application I feel a few things are missing: - your pricing…",[],"unknown","beta",{"praise":83,"objections":99,"requests":143,"questions":163,"alternatives":183},[84,95],{"value":85,"source":69,"evidence":86},"Works well on some images and is easy to use",[87,91],{"id":88,"by":89,"text":90},"5045715","makeee","Did a few tests and it works pretty well! No false positives at least. Any plans to increase the number of objects you can search for at once? Very interested in using this but…",{"id":92,"by":93,"text":94},"5045740","yesimahuman","Works well! Found a few it didn't work on. For example, it didn't detect an airplane in this image (but it's a fighter jet, so maybe not part of the training set):…",{"value":96,"source":69,"evidence":97},"Free tier is appealing",[98],{"id":72,"by":73,"text":74},[100,118,132,136],{"value":101,"source":69,"evidence":102},"Misses many objects such as airplanes and cats",[103,107,110,114],{"id":104,"by":105,"text":106},"5045848","limejuice","Tried detecting Airplanes on this image with 18 airplanes, but it only detected 4 of them.…",{"id":108,"by":105,"text":109},"5045856","Tried to detect Cats in a whole room full of cats, and it detected zero cats. http:\u002F\u002Fenglishrussia.com\u002Fwp-content\u002Fuploads\u002F2007\u002F08\u002F130-cats-1.jpg",{"id":111,"by":112,"text":113},"5045820","mephi5t0","detected 2 planes, there are 7 http:\u002F\u002Fiskin.co.uk\u002Fwallpapers\u002Fimagecache\u002F1280x800\u002Fjet_plane_formation.jpg seems too buggy to pay just yet",{"id":115,"by":116,"text":117},"5046267","cabalamat","It didn't recognise the cat in this picture (http:\u002F\u002Fi.imgur.com\u002FTgFaJh.jpg) so I'm doubtful of it's practicality.",{"value":119,"source":69,"evidence":120},"Reports false positives and wrong counts",[121,124,128],{"id":122,"by":112,"text":123},"5045813","http:\u002F\u002Frumors.automobilemag.com\u002Ffiles\u002F2012\u002F11\u002F2013-VW-Eos-front-three-quarter-with-plane.jpg detected 3 planes... there is only 1 plane and a car",{"id":125,"by":126,"text":127},"5045784","mrtbld","It detects a false positive in this picture for \"full-body person\": http:\u002F\u002Ffarm9.staticflickr.com\u002F8232\u002F8366217251_972624d84b_b.jpg The man is detected, but also a shape above the…",{"id":129,"by":130,"text":131},"5046013","zopticity","Wow, I just tried this image of faces: http:\u002F\u002F3rdarm.biz\u002Fimages\u002F2010\u002F02\u002Ffaces.jpg It got almost all of them but so many errors. It can't detect sheeps either. I was really…",{"value":133,"source":69,"evidence":134},"Video pricing too high",[135],{"id":76,"by":77,"text":78},{"value":137,"source":69,"evidence":138},"Face detection stops at four faces",[139],{"id":140,"by":141,"text":142},"5048068","bsenftner","If you read the description on the site, you'd know that the face detection stops after finding 4 faces.",[144,151,158],{"value":145,"source":69,"evidence":146},"Detect more than four objects at once",[147,148],{"id":88,"by":89,"text":90},{"id":149,"by":11,"text":150},"5045983","Yes! Scanning for more than 4 objects at once (which is currently supported) is something that we definitely want to enable in the future. The only constraint is the number of…",{"value":152,"source":69,"evidence":153},"Confidence scores on detections",[154,155],{"id":76,"by":77,"text":78},{"id":156,"by":11,"text":157},"5045885","Hey steeve, thanks for the really kind feedback! We're aware of the video pricing issue, and it's something that we're thinking hard to come up with a solution to for makers and…",{"value":159,"source":69,"evidence":160},"Video support at lower cost",[161,162],{"id":76,"by":77,"text":78},{"id":156,"by":11,"text":157},[164,172],{"value":165,"source":69,"evidence":166},"Is it built on the Felzenszwalb part-model code",[167,171],{"id":168,"by":169,"text":170},"5046023","dchichkov","It is probably based on \"Object Detection with Discriminatively Trained Part Based Models\"; Pedro F. Felzenszwalb\"... Somebody took http:\u002F\u002Fpeople.cs.uchicago.edu\u002F~rbg\u002Flatent\u002F…",{"id":60,"by":61,"text":62},{"value":173,"source":69,"evidence":174},"Is the Mechanical Turk approach cheaper for accuracy",[175,179],{"id":176,"by":177,"text":178},"5045864","forrestthewoods","Failed completely for me across a half dozen tries. I wonder how cheaply you could get results via Mechanical Turk. I bet you could get much more accurate results for a very low…",{"id":180,"by":181,"text":182},"5045910","aantix","One to two cents a task. Anytime you have a language agnostic task (identifying\u002Fclassifying objects, etc), the tasks can be done very cheaply. Just make sure you do triplicate…",[184,199,206],{"value":185,"source":69,"evidence":186},"OpenCV",[187,191,195],{"id":188,"by":189,"text":190},"5045721","bluishgreen","Can you give me bit more technical background. Tell me how this is better than for eg. out of the box openCV filters.",{"id":192,"by":193,"text":194},"5046022","fchollet","Over a dozen experiments, the recognition rate for faces seems to be about 70%. Example of failure: only 2 faces detected here (in particular NOT the one in focus)…",{"id":196,"by":197,"text":198},"5045808","thechut","Try OpenCV: http:\u002F\u002Fopencv.willowgarage.com\u002Fwiki\u002F",{"value":200,"source":69,"evidence":201},"libccv",[202],{"id":203,"by":204,"text":205},"5046562","liuliu","shamelessly plug: libccv supports REST-ful API in 0.4 version, it is open-source, and free: http:\u002F\u002Flibccv.org\u002Fdoc\u002Fdoc-http\u002F. Trained pedestrian \u002F car \u002F face detectors are included.",{"value":207,"source":69,"evidence":208},"Mechanical Turk",[209],{"id":176,"by":177,"text":178},{"comments":211,"answered":39,"key":212},4,[213,216,217],{"id":214,"text":215},"5045912","Hey everybody, OP here. Thanks for the great feedback! We're really happy that so many people have checked this out. One thing that I want to mention: our service was built…",{"id":156,"text":157},{"id":218,"text":219},"5045953","Hey tunnuz and limejuice, sorry to hear we only picked up on 4 of the planes. We've biased our service towards precision rather than recall; thus, we try to be wrong about…",{"commenters":221,"max_depth":211,"maker_replies":211,"first_comment_mins":222,"lifespan_hours":223},60,21,65,[225,229,233,237,241,244,246,249,252,254,257,260,263,266,267,268,271,274,277,280,283,285,287,289,293,295,298,300,302,304,307,310,312,314,317,318,321,323,326,330,333,334,335,338,340,344,348,352,356,358,361,363,366,370,373,375,378,381,384,387,390,394,397,402,406,410,411,415,419,421,425,429],[226,37,18,227,228,37],"5046004","project\u002Fpositive\u002Fidea","simonsquiff",[230,37,37,231,232,37],"5046031","commenter\u002Fneutral\u002Fnone","opminion",[234,37,36,235,236,37],"5046999","project\u002Fneutral\u002Fperformance","dbaupp",[238,39,39,239,240,37],"5046395","project\u002Fcritical\u002Fperformance","apu",[242,39,36,235,243,37],"5046773","eloisius",[245,39,37,231,240,37],"5050992",[247,39,18,227,248,37],"5049026","EwanG",[250,39,37,251,240,37],"5050987","other\u002Fneutral\u002Fnone",[60,43,39,253,61,37],"project\u002Fcritical\u002Fexecution",[255,43,37,251,256,37],"5046433","DenisM",[258,43,39,239,259,37],"5046655","cynwoody",[261,43,39,239,262,37],"5046774","martininmelb",[264,43,37,231,265,37],"5055221","tripzilch",[176,36,39,239,177,37],[180,36,37,251,181,37],[269,36,37,231,270,37],"5046009","benmccann",[272,36,37,231,273,37],"5046033","trentmb",[275,36,37,251,276,37],"5047304","mbrock",[278,36,37,251,279,37],"5046695","mrgordon",[281,36,39,239,282,37],"5047138","r3m6",[188,211,36,284,189,37],"project\u002Fneutral\u002Fidea",[168,211,36,286,169,37],"project\u002Fneutral\u002Fexecution",[214,18,37,288,11,39],"maker\u002Fneutral\u002Fnone",[290,18,36,291,292,37],"5046020","maker\u002Fneutral\u002Fperformance","rjdagost",[294,18,36,235,265,37],"5055243",[296,30,39,239,297,37],"5045845","senthilnayagam",[192,299,39,239,193,37],7,[301,299,39,239,259,37],"5046711",[140,299,36,303,141,37],"project\u002Fneutral\u002Fdocs",[305,299,39,306,259,37],"5050183","project\u002Fcritical\u002Fdocs",[76,308,36,309,77,37],8,"project\u002Fmixed\u002Fpricing",[156,308,37,311,11,39],"maker\u002Fneutral\u002Fpricing",[104,313,39,239,105,37],9,[315,313,36,303,316,37],"5045871","tunnuz",[218,313,37,288,11,39],[319,313,36,320,316,37],"5046324","maker\u002Fneutral\u002Fdocs",[108,322,39,239,105,37],10,[129,324,36,325,130,37],11,"project\u002Fmixed\u002Fperformance",[327,328,39,239,329,37],"5046027",12,"agotterer",[88,331,18,332,89,37],13,"project\u002Fpositive\u002Fperformance",[125,331,39,239,126,37],[149,331,37,291,11,39],[336,31,39,306,337,37],"5047416","afhof",[115,339,39,239,116,37],15,[341,342,39,239,343,37],"5045746",16,"bbayer",[345,346,39,239,347,37],"5045716",17,"gesman",[349,350,39,239,351,37],"5046166",18,"stakecounter",[353,354,37,251,355,37],"5045995",19,"philhippus",[92,357,36,325,93,37],20,[359,222,36,325,360,37],"5047730","stormen",[203,362,37,251,204,37],22,[364,40,39,239,365,37],"5046075","treelovinhippie",[367,368,36,284,369,37],"5046080",24,"ahc",[371,368,37,251,372,37],"5046150","cookingrobot",[122,374,39,239,112,37],25,[72,376,18,377,73,37],26,"project\u002Fpositive\u002Fpricing",[379,32,39,239,380,37],"5046528","captaincrunch",[382,383,39,239,347,37],"5045742",28,[385,383,36,235,386,37],"5046039","hnriot",[388,383,37,251,389,37],"5046135","keeran",[391,392,36,284,393,37],"5048369",29,"piercebot",[64,395,18,396,65,37],30,"project\u002Fpositive\u002Fnone",[398,399,39,400,401,37],"5047763",31,"project\u002Fcritical\u002Fprivacy","MasterScrat",[403,404,39,239,405,37],"5046153",32,"blaze33",[407,408,36,284,409,37],"5045751",33,"gourneau",[196,408,37,251,197,37],[412,413,36,311,414,37],"5046582",34,"danielharan",[416,417,36,325,418,37],"5046061",35,"misleading_name",[111,420,39,239,112,37],36,[422,423,36,325,424,37],"5045795",37,"pkrein",[426,427,39,239,428,37],"5045825",38,"leoplct",[430,431,18,227,432,37],"5047149",39,"sinzone",[],[435,440,445,450,453,458],{"id":436,"name":437,"points":438,"category":439},"5052155","Teoria.js",77,"languages",{"id":441,"name":442,"points":443,"category":444},"5068547","Super Debugger",75,"dev_tools",{"id":446,"name":447,"points":448,"category":449},"5048009","This Year",70,"productivity",{"id":451,"name":452,"points":448,"category":449},"5066355","Draft",{"id":454,"name":455,"points":456,"category":457},"5057073","analytics-node",66,"data",{"id":459,"name":460,"points":461,"category":457},"5034513","appstorerankings.net",62,false,"v2",{"id":465,"name":5,"line":6,"ai":462,"cat":20,"year":13,"points":14,"comments":15,"rec":27,"s":466,"top":18,"kw":467,"maker":11,"domain":10,"tier":19,"attempts":37,"strip":468,"url":9,"mix":469,"praise":85,"objection":101,"sub":21},5045621,-0.34,[24,25,21,26],"5315101100013311131311311113131313313313",[30,31,30,32],1793990889801]