No. IQ Engines assigns API keys to companies for building apps, and we manage the events (visual query processing, billing, etc) by that single API key. It is up for the app to manage the individual users. In other words, it is an API key per app, not an API key per mobile user.

Yes – there are three settings to consider in making a private visual search system:

(1) Training a private database
You can upload your own set of training labeled images using the Training API (http://developer.iqengines.com/apidoc/current/apis/training-api.html). The default setting is that we match your query images against both your own database and the IQ Engines public image database. However, if you would like to search your images exclusively, we can set your database to “private”, so when a photo with your API key arrives, we only search your private database, and no one else has access to those learned

Yes, we highly recommend before using the Image Recognition Platform to train with known labeled images. That is the only way to get computer vision matches.

There are 2 ways to train the computer vision system:

(1) Training Console (for small image sets)
We’ve created the Training Console to help you try the Training API through an easy-to-use web interface. Use it to quickly train our computer vision with your own database of images and labels.
Link: http://developer.iqengines.com/dashboard/upload/

(2) Training API
For automated training, use the Training API. The Training API allows you to upload images to the IQ

Yes. When you sign up for an API key, you can contact iqengines and ask to have your API key set to computer vision only. This provides more accurate results with proper training. Computer vision will only work for certain classes of objects.

Yes. There are 2 ways we do video search.

(1) Using a mobile device camera to recognize a video
IQ Engines can index a video, and make it visually interactive. So a user can use a mobile phone or tablet camera, point at a TV show or commercial, and the 2nd screen mobile phone or tablet can then take the user to a webpage or video. In other words, IQ Engines API can be used to make video interactive. To do this, we use the Training API to learn a video (typically 2 frames per second), and then the video

Yes, we can recognize barcodes, both UPC and QR codes. The barcode number or QR code text is return through the Results API.

Note: currently the QR Code or Barcode is identified on the server. In the near future, we will also offer client-side QR code and barcode recognition.

Yes, the API performs adult content moderation by default. If we detect an image that contains nudity, violence, or something hateful, then we return in the labels field the string “Objectionable Content”.

We don’t have facial recognition algorithms. Our primarily focus on object/scene recognition for commerce, media, photo tagging, etc. We do have face detection (are there faces) and people detection (are there people) as part of the computer vision suite.

Yes – we offer free pre-build client libraries for Python, Ruby, Node JS, PHP, Java (Android), C# and Objective C (iPhone). You can find these on GitHub: http://github.com/iqengines/

Yes! The image recognition API works for any mobile platform.

For Android, these links will help:
API documentation: http://developer.iqengines.com/
Pre-built client libraries (including Java for Android)

We have many customers that run our image intelligence API on Android apps.

We will release an AR platform very soon. You can view an early prototype of instant recognition (object and location) here: http://vimeo.com/25378976. This is useful for applications that require fast and accurate computer vision for a limited set of objects.

Currently, the response to a visual query is returned in English.

For computer vision training, we have added Unicode support for international character sets (ex. Japanese, Chinese, Korean, etc). This is when images of objects are uploaded via the Training API http://developer.iqengines.com/apidoc/current/apis/training-api.html. So you can train the system to get a foreign language return for computer vision matches from your private training database.

If human crowdsourcing is turned-on for the visual query, the system only supports English for now. Some customers are using the Google Translate API successfully to translate from English to other languages. This works in the

{ ‘data’: { ‘error’: 0 } }

Don’t worry, this is not an error! In fact, it’s a good thing — the
error code of zero is telling you that the image was successfully
received by our server and is being processed. You can check out the
details in our docs:

http://developer.iqengines.com/apidoc/current/apis/query-api.html#return-values

Getting Started. Steps to try IQ Engines Image Recognition Platform

Here is a quick way to Trial the IQ Engines image recognition API.

Step 1. Sign up for an API Key

Step 2. Try the Hello World Demo
Test your API key at the Hello World demo: http://developer.iqengines.com/dashboard/try-it/
This will show you parameters when you post and image and get results.

Step 3. Review the API Documentation
Our full API documentation is here: http://developer.iqengines.com/apidoc/current/

Step 4. Train for Computer Vision
If you have your own database of labeled images, this will greatly improve your search speed and accuracy. With

http://developer.iqengines.com/dashboard/queries/

Once you sign up for an API Key, you get your own Dashboard. You can view your API visual queries and results. IQ Engines is working on more analytics to help you analyze your API history.

With the DELETE function you can remove objects and their associated images from the computer vision database using the API. To ‘edit’ an object’s label, remove it using the DELETE function then upload a new one and apply the correct label. The API documentation includes detailed instructions for deleting objects here: http://bit.ly/ylbP6h.

We use a combination of computer vision and human crowdsourcing to always give results to any visual query (an “any image” platform). Ideally, when an image query is received, the key visual content is identified with our computer vision modules within a few seconds, and that label is returned via the API. If we can not get an accurate result in computer vision, it goes to a suite of crowdsourcing nodes, an innovative network architecture that involves tagging, gaming and social interaction with photos. For IQ Engines, the crowdsourcing technologies are important because they train (a feedback loop) the

Here are two tools to check your API status:

Authentication DeBug: http://developer.iqengines.com/dashboard/debug/authentication/
Hello World Query: http://developer.iqengines.com/dashboard/try-it/

Refer you to example 2 in the docs for the Update API:
http://developer.iqengines.com/apidoc/current/apis/update-api.html
You will see how to handle the case where the Update API times out.

It is not guaranteed that the results for a query will be available before the Update API (though most of the time it is the case, and we are working hard to make sure all queries are processed fast!). Hence, if the Update API times out, the recommended behavior is to start another Update API. The other method to get your results back is to use the callback argument in the

Using your Trial API key, you can train a few hundred images for testing purposes. There are 2 ways to train the computer vision system:

(1) Training Console (for small image sets)
We’ve created the Training Console to help you try the Training API through an easy-to-use web interface. Use it to quickly train our computer vision with your own database of images and labels.
Link: http://developer.iqengines.com/dashboard/upload/

(2) Training API
For automated training, use the Training API. The Training API allows you to upload images to the IQ Engines object search database through a RESTful interface. Once the computer

You can read about authentication here:
http://developer.iqengines.com/apidoc/current/apis/authentication.html#authentication-label

We have created an authentication debug tool which may help.
http://developer.iqengines.com/dashboard/debug/authentication/

For timestamp, you can check the correct timestamp from our Hello World demo page:
http://developer.iqengines.com/dashboard/try-it/
Also, you can run and compare that to the following command in the terminal: date -u +%Y%m%d%H%M%S

IQ Engines’ image recognition platform was developed by 7 PhDs over 3 years, supported by grants from the National Science Foundation. There are several unique features in the platform worth noting:

(1) Mobile and server recognition
IQ Engines image processing can be either on the server (scalable) or the mobile device (fast). A demonstration of both mobile and server object recognition is shown here: http://vimeo.com/25378976.

(2) Any image
When we receive an image query, our computer vision modules identify the key visual content within a few seconds and return relevant keywords via the API. However, if we cannot

Make sure you do not use images that are too small. We recommend images from 300-600px in size; this is ideal for speed and accuracy. Do not submit images smaller than 200px.

IQ Engines uses computer vision for object recognition, but we resort to human crowdsourcing (image tagging) if computer vision is not successful. We built our crowdsourcing network and techniques to optimize speed and performance. In the most simple sense, IQ Engines crowdsourcing is humans doing object recognition. It’s not perfect, but we think human crowdsourcing will get better soon, with “expert tagging” and “social tagging” projects in the pipeline, that will result in more detailed and accurate tags.

One special feature IQ Engines offers is a “custom question” for API users that rely on crowdsourcing. That means, along with the

oMoby (www.omoby.com is a reference iPhone app that uses IQ Engines visual search platform. You can actually plug your API Key into the oMoby app and test your private trained computer vision images.

Our address is:
IQ Engines
2501 9th Street, Suite 100
Berkeley, CA 94710

Our phone number is: (510) 592-4731

To balance speed and reliability, we recommend you train and search with images that are 400-500px (minimum 300px).

Pricing can be found on the IQ Engines website. http://www.iqengines.com/pricing/.

The API sign up is here: http://developer.iqengines.com/

The API document is here: http://developer.iqengines.com/apidoc/current/