We usually "region propose" and crop to a certain area (in this case the face area, usually at 256x256) then transform to align eye areas before passing to training. This is to standardize the data beforehand. I'm not sure if this lib does region proposal but you can easily write a pre-processor with openCV face plugins to identify face regions (if any, maybe your training image is a landscape not a face!) for cropping.
Yes, the processing pipeline first does face detection and a simple transformation to normalize all faces to 96x96 RGB pixels. Then each face is passed into the neural network to get a 128 dimensional representation on the unit hypersphere.
For a landscape, face detection would probably not find any faces and the neural network wouldn't be called.
And an image with multiple people will have many outputs: the bounding boxes of faces and associated representations.
Thank you for applying for the Y Combinator Fellowship. We’re sorry to say that your startup was not selected for an interview. We think your startup would be more appropriate for YC’s Winter 2016 batch, and we encourage you to apply.
The startups we have invited to YC Fellowship interviews are primarily founders in the idea or prototype stage, and most are first-time founders who haven’t previously raised funding. We're asking almost all companies that are further along or with more experienced founders to apply to YC winter batch.
Applications open the week of August 24th at apply.ycombinator.com. We sincerely hope you will apply!
Yeah, and even for non-clothing things, you might be able to make a really cool system that retailers can sign up for to list their inventory with you, so that people have an awesome online way to find things they'd otherwise just buy on Amazon in a brick and mortar store, they can get to today.
Build out social widgets and interactive map themes for our mobile front-end and our bubble builder CMS geo plotting admin interface.
---- Knowledge of (or eagerness to learn): Angular.js (Javascript), CSS3, HTML5, PhoneGap, Git
---- Nice to have: Leaflet.js
======= About Us ========
Our first product Bubbl.li organizes the information relevant to you and makes it easily accessible when you need it. Our technology creates “databubbles” that are pinned to specific places and times. Users inside a databubble can access available information and have it personalized to their needs.
=========================
Interface Foundry is a Public Benefit Corporation. We believe in open source and transparency.
Our culture embraces diversity, we value a collaborative environment with lots of challenges that require creative solutions. We like walking-breaks, snacks and cats. We’re flexible about time as long as you get work done.
You’ll be joining us in our office in Flatiron, Manhattan NYC. When employees join full time, they receive:
• Stock options with vesting interest
• Health insurance
• Flexible scheduling
======= Contact ========
Interested? Send us a note:
hello@interfacefoundry.com
Develop our platform's back-end automated systems for API consolidation, widget deployment, and communication with 3rd party connected devices. Also, all levels of security and database optimization.
• All early employees get both stock options and vesting interest when they join full time, along with health insurance.
• Office located in downtown SoHo, Manhattan, near lots of great food in Chinatown.
//=== About Us ====//
Interface Foundry's mission is to index the physical world, allowing people to move seamlessly from one area or timeframe to another, interacting with their environment where information (apps, widgets, connected devices, data cards, localized maps) is absorbed into personal devices securely.
Our platform lets you browse & build databubble micro-worlds -- data pinned to a specific location or time, consolidating apps and other widgets into a single context-aware interface.
We just closed our first round of seed/angel financing, raised over 100K and looking to grow and scale quickly. We have over 20,000 users, deployed worldwide everywhere from Berlin to Barcelona, Detroit to Oakland and NYC.
Our platform is open source and open API. We believe strongly in transparency and accountability, and have benefit-corporation status.
//=== Apply ===//
If you’re interested, send us a note: hello@interfacefoundry.com
The phones don't just go away, they're get passed down.