{"id":342384,"date":"2017-05-31T22:00:10","date_gmt":"2017-05-31T14:00:10","guid":{"rendered":"http:\/\/techcrunch.com\/?p=1497508"},"modified":"2017-05-31T22:00:10","modified_gmt":"2017-05-31T14:00:10","slug":"miso-happy-is-a-3d-selfie-app-that-rocks-the-uncanny-valley","status":"publish","type":"post","link":"https:\/\/people.utm.my\/asmawisham\/miso-happy-is-a-3d-selfie-app-that-rocks-the-uncanny-valley\/","title":{"rendered":"Miso Happy is a 3D selfie app that rocks the uncanny valley"},"content":{"rendered":"

<\/p>\n

Move over duckface. Shove off rainbow spew. There\u2019s a new selfie game in town and it wants to turn your 2D visage into uncannily accurate 3D \u2014 then splice that axiomatic piece of you into animated cartoon suits that rock out, strut their disco stuff, skateboard, look coy, relaxed or laugh manically.<\/p>\n

Think: realistic-looking stickers that twist and twirl. If that idea doesn\u2019t immediately thunder with resonating intent through your mental social feeds this app is definitely not for you. It\u2019s definitely not for me \u2014 but selfie is as selfie does. And those that selfie rarely stop. They just need new twists to keep upping their gaze-game.<\/p>\n

And so we arrive at\u00a0Miso Happy<\/a>. Terrible name, terribly creepy results\u2026 Expect this 3D selfie maker app to either have you shrieking with delight \u2014 or plunged deep into the uncanny valley reaching for the \u2018delete face\u2019 button with a horrified expression on said face.<\/p>\n

The Finland-based startup behind the app, Futurefly<\/a>, pitched it to TC as an \u201caugmented reality sticker faceapp<\/a>\u201c. Which is a little linguistically hacked together but did get the message across. On the App Store it\u2019s billed as \u201cyour 3D selfie effect\u201d.<\/p>\n

The app\u2019s (word-punning) name suggests the team reckons it will do well in Japan. As does the fact one of the first clutch of cartoon suits is described as a \u201cshy Anime schoolgirl\u201d. There\u2019s also a sumo wrestler suit, along with a skater kid, a snowboarder, the above pictured glam rocker, a disco dancer, an angel, a couple of exhausted looking beach bums, and more.<\/p>\n

Miso Happy is AR because once you\u2019ve snapped a photo and generated your creepily realistic (\u2018uhhh, that really is my face<\/em> licking that guitar\u2026 O___o \u2018) 3D selfie, and dressed it in the animated suit of your choice, it then appears doing its routine in the camera view \u2014 waiting to be video overlaid against the real world backdrop of your choice.<\/p>\n

You can also add text and emoji overlays to your composition, pinching to zoom any other elements up or down in size as you like.<\/p>\n

\"\"<\/a><\/p>\n

Then, once you\u2019re content with your content it can be shared to social feeds like WhatsApp and Facebook as a video (or a GIF), or saved locally on your iPhone if you don\u2019t want to immediately spam your social graph with all your 3D action selfies.<\/p>\n

There\u2019s a pretty small selection of cartoon body suits to choose from at this launch point, but the team says it will \u201cregularly\u201d be adding new ones, and that there will be no need to update the app to get them. This is also Futurefly\u2019s monetization route \u2014 with the plan being to offer twelve free 3D animated stickers per day, plus three locked stickers which can be unlocked by watching one video ad.<\/p>\n

The app also lets you store different selfies, so you can keep a library of friends\u2019 selfie \u2018face masks\u2019 to play around with. Assuming that\u2019s the kind of friend you are.<\/p>\n

Computer vision<\/h2>\n

What\u2019s the core technology here? Computer vision of course \u2014 a very hot area for tech activity (and acquisitions<\/a>) these days. For example Snapchat snapped up another European startup offering 3D selfies (Seene<\/a>) last year. (Likely feeding some of that tech into its World Lenses launch<\/a>, which lets its users add AR 3D objects to their Snaps.)<\/p>\n

\u201cThe core face reconstruction technology combines computer vision and computer graphics to get the final results,\u201d says Oskari H\u00e4kkinen, founder & chief product officer at Futurefly, explaining how Miso Happy achieves its uncanny 3D effects (and pictured below putting his own face through the process). \u201cWe use landmark detection to detect the eyes, mouth, nose and the contour of the face. This is the same technology which is used in apps like Snapchat and Facebook Messenger when applying face filters.\u201d<\/p>\n

This translates in app to a marked face template at the point of selfie taking which encourages the user to correctly position their face within the guide lines. They also suggest using a white\/light background for best results.<\/p>\n

\"\"<\/a><\/p>\n

\u201cAfter the user snaps their selfie with Miso<\/span> Happy, we approximate the camera angle and fit the users landmark data into a generic face morphable model, which returns us the 3D vertices. Next we project the selfie pixels into a texture map and onto the 3D model. Finally we post-process the texture illumination and remove any strong specularities,\u201d he adds.<\/p>\n

H\u00e4kkinen says the tech was challenging to develop, although from the user point of view it\u2019s disturbingly quick and easy to turn a selfie into facial 3D. The processing is done locally on the device to ensure there\u2019s as little delay as possible while you wait for your selfie to be reformed into digital 3D.\u00a0The app is optimized for iPhone 5 and above and iOS version 8.1. (An Android version is scheduled for later this year.)<\/p>\n

Another challenge during development was creating a realistic enough 3D likeness that users wanted to share the results.<\/p>\n

\u201cWe had our initial implementation of the technology in September last year, which we tested with college students. At first, they were stoked, laughed a lot and said it was awesome. However, when it came to sharing they actually didn\u2019t want to share it with friends because it didn\u2019t look enough like them,\u201d H\u00e4kkinen tells TechCrunch.<\/p>\n

\u201cAt the time the method we were using was to project a texture onto a morphable avatar that users had to morph themselves. But after time we realized that people actually don\u2019t know the exact shape of their faces, and therefore couldn\u2019t really morph the avatar accurately enough to be left satisfied with the results. With that, we knew we had to automate that process, else users would always feel the results were \u2018slightly off\u2019. The new tech has an automated morphable face model that adjusts with the retrieved face data.\u201d<\/p>\n

So while the results are still uncanny valley levels of creepy IMO, clearly they were a lot MOAR uncanny before.<\/p>\n

The team is also working on adding\u00a0further post-processing features such as \u201cbeautify filters\u201d, according to H\u00e4kkinen. And is developing augmented reality surface detection to enable the 3D characters to \u201csnap to different surfaces within the world\u201d. At the moment the animated figures just float in the middle of the screen or wherever you generally position them.<\/p>\n

Who and\/or what is Miso Happy for? Clearly it\u2019s just for fun, and aimed squarely at Generation Snap\/selfie. The team\u2019s longer term gambit is perhaps to be (ho-ho!) snapped up by a larger entity that appreciates the visual expertise they are now demoing in the wild.<\/p>\n

\u201cWe just hope people will jump on this, create fun content and tell us what kind of new animations they want. Content is pretty easy now for us deliver, and we can deliver it pretty fast and through the backend, so we hope people will come back daily to check out the new content daily,\u201d\u00a0says H\u00e4kkinen.<\/p>\n

\u201cFor now it\u2019s just a content creation tool, much like Snapchat and Messenger filters, but with a totally new and unique twist. I guess nobody needs faceapp<\/a> to make museum paintings smile, but it\u2019s fun as hell, so who cares, right?\u201d he adds.<\/p>\n

Asked whether they\u2019d be open to acquisition approaches from the right kind of buyer, his response can be summed up as \u2018definitely maybe\u2019. \u201cWell, it\u2019s a mixed bag really, we love to be independent, but sometimes partnering up can help your products reach the largest possible audience, which is why we get up each morning. That said, we are currently in the process of raising our next round with the intention to grow our user base,\u201d he says.<\/p>\n

Futurefly, which is headquartered in Helsinki and was founded in 2014, has raised $3M to date, from private Silicon Valley investors who were early investors in Snapchat, WhatsApp, Twitter and Facebook \u2014 so it certainly has lines in to the biggest social platforms around.<\/p>\n

Other investors include Arielle Zuckerberg, Betaworks, Broadway Video Ventures, Vision + and Risto Siilasmaa (chairman of Nokia).<\/p>\n

This is one of a handful of apps the team has made. One of its earlier products, Rawr Messenger<\/a>, which sticks 3D avatars into a messaging space, has more than a million users, according to\u00a0H\u00e4kkinen. He doesn\u2019t break out user numbers for another of its apps\u00a0EatMessage<\/a>, which is billed as the\u00a0\u201cfirst ever selfie game where you use your mouth as the controller\u201d.<\/p>\n

Futurefly is surely hoping Miso Happy will be the viral hit that helps it make more of a mainstream splash.<\/p>\n","protected":false},"excerpt":{"rendered":"

\"\" Move over duckface. Shove off rainbow spew. There’s a new selfie game in town and it wants to turn your 2D visage into uncannily accurate 3D — then splice that axiomatic piece of you into animated cartoon suits that rock out, strut their disco stuff, skateboard, look coy, relaxed or laugh manically. Read More<\/a><\/p>\n

\n<\/img><\/a> <\/img><\/a> <\/img><\/a> <\/img><\/a> <\/img><\/a> <\/img><\/a>\n<\/div>\n

\"\"\/ <\/p>\n","protected":false},"author":5817,"featured_media":0,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_et_pb_use_builder":"","_et_pb_old_content":"","_et_gb_content_width":"","footnotes":""},"categories":[25],"tags":[59,66,26],"class_list":["post-342384","post","type-post","status-publish","format-standard","hentry","category-technology","tag-media","tag-techcrunch","tag-technology"],"_links":{"self":[{"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/posts\/342384","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/users\/5817"}],"replies":[{"embeddable":true,"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/comments?post=342384"}],"version-history":[{"count":0,"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/posts\/342384\/revisions"}],"wp:attachment":[{"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/media?parent=342384"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/categories?post=342384"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/tags?post=342384"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}