Analyze Images and Collect Data - Repost - open to bidding

We have a data collection project in progress and we are looking for two additional service providers to work with our team. Basically, we require a competent individual to review a sequence of images and also capture place names into a provided web application. This includes the name of place/business and other information from within the images. You will first be provided with more images with annotations for training. Sample images and data provided here:- [login to view URL] On average there are 5 businesses in every 4 images. To start, we will provide you with 100 images to process and then we will assess the quality of your work. Quality is key to our business. Once satisfied with the work we will continue with you for the duration of our project which takes us to Oct/Nov. What we specifically require:- 1. To review the high quality jpg images weekly. Each image is approximately 4-6MB. These will be made available via .rar files in Dropbox. 2. To input data of the place/business/businesses captured in each image accurately into the web-based information system provided. 3. To review images to discern the name of respective business/place, select the type of business/place/activity from a drop list, contact information, days and hours of operations, payments accepted and any other important information as defined in the web application, where available. An image may have more than one business/place. 4. Some images in a sequence may have the same information but each may be able to give additional information on the place/business. In this case you wouldn’t be required to enter the same info twice. 5. To maintain absolute confidentiality. You will need:- 1. Reliable Broadband Internet 2. An image reader that allows basic zooming 3. Understand Basic English language 4. Pay attention to details. 5. Be able to look at the image and ascertain the nature of a business /place if the name doesn’t suggest. The submitted data will be evaluated for quality. 33% of the images will be randomly chosen and assessed against what you have captured and we expect no more that 1% of the captured data to have errors.