• Time-saving tips
• Uploading images
• Automatic classification
• Main keywording page: My Images
• Commercial Vocabulary
• Hotlists
• Shared Hotlists
• Controlled Vocabulary
• Exporting keywords
• Agency export
• XMP and IPTC embedding
• Using the Adobe Bridge plugin
• Downloading the Controlled Vocabulary

To get started with keywording as fast as possible, first type in the caption, then click the magnifying glass icon next to the caption box. Annotator will analyze the linguistic structure of your caption, pick out the salient terms, and give you a variety of keyword recommendations to save you time.

If you have captions embedded as XMP or IPTC headers in your uploaded images, Annotator imports these automatically. Then you can start keywording just by clicking the caption search button.

To add one or more keywords rapidly without searching and without using the mouse, you can type Ctrl+Enter after typing each keyword into the Search box.

You can batch-delete a keyword from multiple images. Simply select the images and click the delete icon next to the keyword. If your selection also includes images that don't have that keyword, then the keyword won't appear (because the set is the 'intersection'). In this case, you can use this short-cut: batch-add the keyword again, to your entire selection. Then you can delete it.

The first step in using Annotator is to upload your images in the Upload page. You may upload either a single image or a Zip file containing a batch of images. Each image you upload should be about 1.5 MB uncompressed ('open file size' in Photoshop), or about 850 pixels on the longest side. Images of this resolution are about 100-300 kB on disk when compressed as JPEGs, depending on the level of detail in the image and the compression level.

You may choose which folders you upload your images into, and you may choose to preserve any existing folder structure in a Zip file. (To create or change folders, use the My Images page.)

Uploaded images then go through imense®'s processing pipeline to classify some of their content automatically. This requires a minimum resolution of 0.2 megapixels (at least 370 pixels on the shortest side) and takes up to 10 seconds per image. You can see the list of files still in the processing pipeline on the Upload page. When all files have been processed, proceed to the My Images page.

(Images do not need to be uploaded when using the Bridge plugin.)

imense® uses computer vision technology to label some of the content in your images automatically, including how many people are in an image, their age groups, sex, and ethnicity; whether the image is a cut-out; whether the image contains copy space; whether the image is color or black & white; an illustration, macro shot, or general photograph; and some basic keywords describing the scene. We are continually expanding our range of classifiers. Our ultimate aim is to take away as much of the drudgery of keywording as possible.

The My Images page is the central control panel where you will do most of your keywording. Here you can add keywords and other annotations to your images, either in bulk or one at a time.

You may organize your images into folders for convenience. To create a folder, click the + button next to a folder name. To rename a folder, double-click its name. As a safety feature, deleting a folder is only possible if you have first deleted all the images and sub-folders that it contained.

An image can be selected for keywording by clicking on it. Clicking on the larger thumbnail that appears will zoom the image to its maximum size in the browser window.

Images can be edited in bulk by selecting the check-boxes under multiple images, or for a whole page or folder at once by clicking "Select page" or "Select folder". The keyword list (at right) when bulk-editing contains only those keywords present in all of the selected images. Likewise for the release status, license type (RM/RF), etc. So, for example, to designate 15 images as being cutouts, just select them all and click the "cutout" check-box.

To add keywords to your images, use the third column. The most powerful feature is the Search box at the top, which finds related terms from three sources: your Hotlists (including Shared Hotlists), the Commercial Vocabulary, and the Controlled Vocabulary. You can then add any of these suggested terms to your selected images just by clicking them.

As you click these terms, these appear in the Keywords box in the right-most column. You can do further searches on these terms by clicking the magnifying glass icon next to each keyword as it appears. To delete any unwanted keywords, click the red x button.

Once you have your keywords, you can re-sort the keywords by dragging them up and down with the mouse. This is useful if you submit your images to agencies like Alamy that require keywords to be sorted in priority order. Tip: If your browser isn't letting you drag a keyword to the very top of the list, simply drag the top keyword down instead.

The Commercial Vocabulary is a powerful time-saving tool that gives you a wide range of keyword suggestions. For example, if your image features fruit, type "fruit" into the Search box. Then the Commercial Vocabulary gives you a list of suggested keywords related to fruit, like "apple", "berry", as well as more descriptive terms like "flavor", "nutrition", "citrus". The Controlled Vocabulary also gives you a more extensive list of fruits, divided into groups, like Amazon fruits and different kinds of berries.

Your hotlists are personalized lists of keywords that can greatly speed up repetitive keywording. To make the most of them, use them to define shortcuts for 'bags of keywords' reflecting your own artistic interests. For example, if you regularly photograph animals in the wild, consider making hotlists for concepts that commonly appear in your material -- such as these:

• cub: animal, babies, baby, cute, cub, cubs, feeding, maternal, nature, nurturing, playful, playing, protected, suckling, young, wild, wildlife
• hunt: animal, brutal, feed, feeding, food, hunt, hunting, predator, prey, savage, survival, victim, voracious, wild, wildlife

This way your hotlists will be complementary to the Commercial Vocabulary, which covers a broad range of general material.

To create and edit your hotlists, click on the Hotlists tab at the top of the screen. The page will appear like this:

In the "My Images" and "Keywords" pages, after searching for a term that appears in one of your hotlists (or shared hotlists), you can quickly add all the keywords in that hotlist by clicking the hotlist title (in bold). The title itself is not added. This allows you to define hotlists with titles that aren't themselves valid terms. For example, "wildlife_beauty" and "wildlife_savagery" could reflect different emphases for two hotlists containing wildlife terms.

You can share hotlists with other users. Shared hotlists appear in the search box on the Hotlists page, on the Keywords page, and on the My Images page. To enable this feature, select the "Share hotlists" checkbox on the Settings page:

Enabling this feature shares any new hotlists you create with other users and allows you access to other users' hotlists. Hotlists are shared by default.

If you wish to keep your hotlists private, you may deselect the "Share hotlists" option at any time. If you do this, new hotlists you create will be hidden from other users, but you will also lose access to the pool of shared hotlists.

The Controlled Vocabulary has a number of smart features. One is that it knows about synonyms – different words that have identical or very similar meanings. These appear together in the same gray box when you hover the mouse over them. By default, when you click one to add it as a keyword, the others are automatically added too. You can turn off this feature by clearing the "Auto-expand synonyms" check-box in the Settings page.

Another smart feature of the controlled vocabulary is that you can automatically add parent keywords called hypernyms. For example, if you click "puma" in the controlled vocabulary, you also get the keywords "big cat" and "cat". If you click "figure skating", you also get "ice skating" and "skating". To turn this feature off, clear the "Auto-expand hypernyms" check-box in the Settings page.

A third smart feature of the controlled vocabulary is that it knows about spelling variants for different English locales (US and UK). Your locale setting is in the Settings page. You may also choose to show either US or UK spellings, or both if you want to keyword your images for international buyers.

After keywording your images, you can export your keywords and other annotations as an spreadsheet (in Excel XLS format) or as a CSV file. You can open XLS files in either Excel, OpenOffice Calc, or iWork Numbers to check them before sending to your agency.

You can choose whether to use a comma or some other character to separate your keywords in the Settings page.

For most users, an Excel XLS file is the easiest spreadsheet format. If your agency requires a CSV file instead, here are some tips for importing CSV files into spreadsheet software:

Importing CSV into Excel Instead of double-clicking on the CSV file, load Excel first and then choose Data | "Get External Data from Text". This avoids a bug in some versions of Excel that skips the Text Import wizard when importing a CSV file and assumes default settings that are incorrect. Also, Excel has trouble reading CSV files when any images a Caption or Description field with multiple lines (line breaks). In this case, export your metadata in XLS format instead or use OpenOffice.

Importing CSV into OpenOffice Calc If you have non-English characters in your metadata or filenames, you need to tell OpenOffice explicitly that the exported CSV file has a "Unicode (UTF-8)" text encoding, by choosing this from the Text Import dialog box. This preserves the non-English characters correctly.

Photographers who submit images to Age, Alamy, Corbis, or Getty can save time with special export formats for these agencies. This automates much of the tedious process of making sure your metadata is ready for submission.

For all agencies, Annotator automatically uses the appropriate terminology. For example, some agencies require "Color / B&W" versus "Yes / No" in the Color field, or "Nobody" versus "No people", or release written as "NR" versus "NA" versus "No".

If you select Alamy in the "Export format" box on the Export page, Annotator also splits your keywords into three fields: Essential, Main, Comprehensive, obeying the Alamy character limits for these fields. First make sure your keywords are sorted in priority order. (You can do this on the My Images page by dragging them up and down with the mouse.) Check your metadata spreadsheet for errors; if it looks correct, email it to memberservices@alamy.com. Alamy will then apply the new metadata to your images on the live Alamy site.

On the "Settings" page, you can either choose a keyword separator manually for your spreadsheets or CSV files, or you can choose 'auto'. If you choose 'auto', Annotator will use the correct separator specified by each agency: pipe character (|) for Corbis, space for Alamy, or comma for the other agencies.

By the way, non-space separators will display with a trailing space in the Export page preview, even if you don't ask for a space. Don't worry about this — this is a workaround for web browsers' limitations with line breaks. Your exported file will be fine.

Annotator can now embed XMP and IPTC metadata directly inside JPEG and TIFF files on your computer. This requires Sun Java v5 (also called v1.5) or higher. To get Java or see if your computer has Java, click here.

Normally browsers block access to your files. For XMP / IPTC embedding to work, you must grant Annotator elevated security privileges by clicking "Run" (or "Trust" under OS X) when your browser asks you whether to trust the certificate from annotator.imense.com. This will allow Annotator to write metadata into the images on your computer.

To avoid having to click "Run" or "Trust" every time you use Annotator, check the "Always trust content from this publisher" box under Windows. Or, under OS X, click the "Show Certificate" button and check the "Always trust annotator.imense.com" box.

Warning! Always keep a backup of your high-res assets! If your computer loses power or something else goes wrong while writing metadata into an image, that image could be destroyed.

The Annotator plugin for Adobe Bridge (currently in beta) allows Annotator subscribers to apply keywords to their images directly from within Adobe Bridge, as a simplified alternative to the web interface for keywording. The Bridge plugin supports XMP embedding but does not support agency exports or other advanced features from the Annotator web platform.

If you use Adobe Bridge CS4 and wish to try the Annotator plugin for Bridge, download the plugin from the Settings page and install it by double-clicking the .mxp install file. Then follow the prompts from the Adobe Extension Manager to install the Annotator plugin.

After installing the plugin, start Adobe Bridge and go to Tools | imense® Annotator. Enter your subscription details at the login screen. Then you will find the usual Hotlist, Commercial and Controlled Vocabularies to use to directly keyword selected images within the Adobe Bridge application. Adobe Bridge embeds these keywords as XMP metadata in your image files.

Note for Windows users: If you use Windows Vista or Windows 7, the first time you run the Imense plugin for Bridge you must run Adobe Bridge as an administrator. To do this, right-click the Adobe Bridge icon and select "Run as administrator". This works around a bug in Adobe Bridge on these platforms.

You can download the imense® Controlled Vocabulary for your own use if you wish, subject to our terms and conditions. To do so, click the download links in the Settings page.