Exploring Excire

 I had gotten interested in Excire at the Visual Story Tellers on-line conference. It seemed to offer some useful features that might help me tackle the process of working through a lot of old family photos, slides, negatives and albums. Well, it didn’t inspire me to start that project. However, it is an amazing program.

The clearest use case for excire is during the import/ingestion of photos. It was faster than lightroom, but it does have to create its own database (like the lightroom catalogue) and it uses either a jpeg or the jpeg built into any RAW file.

The fact that it can write this keywording and your ranking to a sidecar file as you import is a really massive benefit.

I still think photo mechanic lets you get through the import a lot quicker, but I’m seldom adding keywords. Once you have entered custom keywords both do a great job of finding your photos.

The second great application is for searching through a large collection, where its hierarchical keywords strategy can help you find a lot of photos quickly within a larger collection. It also keywords several special terms related to aspects of the photograph, not things buried in the exif data (like lightroom) but general things like brightness, leading lines or silhouette.

Its biggest downfall for my “family history” project was that although you can put in customised to identify people and grouping people “like this” it misses enough to make it unreliable, as a fall-back organizer.  By comparison google photos gives me a slightly better strike rate once I identify people (or alternatively say not matching). Picasa (although no longer supported by google) still appears to me to do the best job and It works on photos locally.  I want to use the best tool, with assistance from picasa and photo mechanic for now.

So my conclusion is I’m not buying excire (just now). It is an impressive program and leads the way for AI classification and assistance in finding photos. It's definitely something to keep an eye on. There is a free seven-day trial if you are interested.           

Automatic Classifying the Subjects in your Photos

 One holy grail in the AI applications in Photography (or any visual art) has been the potential ability to automatically see into the binary files (that are photographic images) and use that in searches. For a decade or so google has claimed a lot of progress, starting with the really useful ability to find cats in photos. Ok I’m being a cynical but I have been tracking the claims in google photos, auto labeling things and places. It can look through the great dumps of photos its backup & sync tools is daily uploading. Are things improving? You can enter a term and be “surprised” (on the first couple of occasions) that it will find some photo you have forgotten about. It does also pull up a few clearly bad matches. What I have noticed is the really obvious things google misses in many selections. So I’m really not seeing any significant improvement.

For a while at least flickr had a feature that would add tags at the time you uploaded your photo. If you bothered to check you were able to delete incorrect terms of add your own.  This seem a better approach to me but alas it seems to have fallen by the way side now.

Moving forward I’ve now trying out a slightly new alternative Excire (it has been around as a plug-in for lightroom as Excire Search for a while) which primarily groups and classifies your images to create a hierarchical set of keywords. It can then write these keywords to sidecar files, which will then allow this information to percolate into any other software that reads this format. I’m only using the trial version which unfortunately doesn’t give me the ability to write these files, so I have check it yet. Its classification strikes me as both friendlier to use and possibly better than either of the above.

It also have the benefit that you see the grouping visually and they are laid out from best match at the top to just possible at the bottom. Further you don’t have to change modes (eg. go to new screen to make corrections). You just select the image click on it and you can make changes to the keywords on the left tab or albums on the right.

It strikes me as perfect to be run as you first upload/ingest your photos. Helping you to rank and cull but giving the incentive to better keywording of your images. I already have photo mechanic (no AI features as yet) which I love and use (but not so much for keywording). So I’m Impressed but not ready to jump ship just yet, for keyword Shangri-la.  However that time is much getting closer.

Flickr’s Autotagging

I have been impressed with flickr’s autotagging of photos, It is much more precise than google photos’s search groupings. My thought was always that the google photo system should give the user the ability to flag their photos and say for example this is not a cat However I could never find such and option. Up to now I had not felt inclined to try and correct flirks system, which occurs as you load the photo, Today when I went to remove a couple of tags incorrectly allocated to this photo of a ring tail possum, which flickr had tag as a cat and a pet, and the lattice fencing it is peering through as a photo border, I found I could not delete the autotags. So flickr is also missing the opportunity to have users help improve their classification.

 Flickr was a very earlier introducer of keyword tagging and you can have up to 75 tags, This helps both yourself and others find the themes in photos. The introduction of autotagging happened last yeast and these tags are differentiates by being in a white box with a grey outline. Whereas the tags you add are shown in a grey box. To remove a tag you just hover the mouse over it and a X shows up on the right hand side, clicking on it then deletes the tag. The issue I discovered today that this  while this works on your tags (the grey boxed tags) it does not work on the the autotags (white boxed tags)

PS: Just a day after I posted the above obserbvation the ability to deleted auto tags has returned. It probably only went away last week. However judging my the help forum thread a lot of other users noticed it as well. I appluad the flickr team for respondng quickly. Unfortunate the forum post associate this with a possible drop in the performance of correctly classifing photo, but I must admit I find it hard to believe that this might be cased by users deleting miss classified photo, I think that should improve classification.

Virtual Albums

One aspect of digital photo management that I like is the ability to create virtual albums (collection in lightroom).  These are not duplicate copies of the photos just different views that show only the photos selected without having to move to the folder/directory in which it is stored.  The photos remain in the original folder on the computer, there is just an additional link entry in your album to find that photo and display it.  It is the perfect way to store and show only a best photos or photos on a given theme. They are a great way to organise your photo collection and probably deserves a more detail post in the future.

The downside using an album structure in current software is that it can be very hard to transfer this structure to other computers and/or software.  Each package has the unique way to maintain the album contained within its code.  A few packages will let you exports and upload the contents of albums but fall short of being able to export the album structure,    For example Lightroom does allow their collections to be exported as lightroom catalogues, then reimport this catalogue into lightroom on a different computer.  The process is simple enough but is better described by Steve’s Digicam post. Another examples is the original Picasa and Picasa Web Albums (the on-line side, which is now google+ photos) which lets you maintain matching Albums on-line as on your computer and synch the two.

Use Album Keywords as Metadata

A neat way I have to be able to exchange the the organization you have established with your albums and then be able to exchange this between packages or onto online services that support the Album concept, is to code all the contents of the album (or collection) with a unique Keyword, tag or Hashtag. I was for a time using the ~collection_  prefix to identify this organization metadata by I have abbreviated this to ~alb_ (for Album). Of a  photo or batch of photos is transfered it is usually easy to search for the new metadata, Tags or Hashtags and create a new album or similar grouping from the search results.

To finish here is a great post by +Will James, on how to reorganize your albums in google+/picasa web albums.

Keywording your photos

eywording (or tagging) your photos, sounds simple enough, but if you haven’t already started the task could have already got away from you.For someone taking and preparing a lot of stock photographs, some form keywording is probably essential but the time required and benefits trade-off for the rest of us will not be so clear. I take a wide variety of photos anyway, many are just for personal use, for painting reference, for use in collage or panoramic and a lot are family etc. My most important task is to to make these separations in my collection, I could do that physically by storing them in separate folders or even different hard disks. However, using key words and simple filters with do the same job from a unified collection (see my strategy 1 below). Designing a perfect classification or keywording system is a really broad topic but a good start (I’ve been using this approach for a while, but I have to admit not everything is keyworded yet) is to use a hierarchy based on the classic Who, What, Where, When, Why and How.  You probably won’t uses these categories as keywords themselves but they should help guide you as the most likely way to “find” an individual photo in the future. The Who? is pretty straight forward, and there are a number of ways to tackle this but a person’s name is usually fine, but this can be a lot or work for big groups/events, Lightroom affectionardos might like Gerard Murph’s method of custom keyword sets. The What? is very obvious you should describe what is the subject of yoour photo. This is likely to be the biggest part of your keyword list and think of this as a hierarchy and including each term for each level of the hierachy (see my strategy2). I remember hearing Varina Patel discussing this on a podcast and giving good advice on keywording by theme particular if you intend to submit photos to stock agencies. Several of the keywords like the When? and Where? will be best to be automatically read from the Cameras EXIF files, or batch such as GPS coordinates in a merge of .gpx files. There is also a lot of How? details also available in the EXIF data, but a few special keywords to relate to technique won’t go astray. It is the Why? that most people probably overlook, but the reason you took the photo in the first place is likely to be the quickest way to find an image!

There is a massive trade of here, adding keywords will undoubtedly make photos easier to find, but how much easier and how munch time is saved is pretty much an unknown at the time you are supposed to be doing the keywording. Against that is the amount of work you have to put in tagging and classifying your photos.In particular everyone will tell you you need to do all this extra work up front. Otherwise it becomes too large a task to tackle. So how do you get started, here are some good strategies I have found

1. Have a few (less than 12) broad categories (eg. Portrait, Landscape, Family, Event) so one of which at least is used for every photo, and applied as or soon after you upload your photo. These should be based on the themes and types of Photograph you like to take. If one category gets particularly top heavy consider splitting it into some sub-categories and always add those as a second keyword.
2. Aim to use at least 3 keywords for every photo that correspond to different levels in your classification Hierarchy (Its simple maths the more levels of categories you use the easier it is to find a specific photo). Yet be constrained 7 or more keywords may not give you much advantage in the end.
3. When creating new keywords, avoid plurals, compound hyphenated terms (it is better just to use the two words as individual tags) and ambiguous terms. Watch out for misspelling.
4. Use albums (or Collections in Lightroom) to subset important groups when attempting to batch recode keywords into exiting photo collections. The big advantage is you can “see” the collection together. Taking time to make special collections and batch keywording them at the same time is the quickest way to keyword unorganised photo archives and avoiding this being a chore.
5. Tagging your photos with your keywords, as well as embedding them in your EXIF metedata when you post them on social media for public view if a great way to get extra traction  Unfortunately once again there is no uniform of standard way to do this, but it is worth investigating your favourite publishing places and services (eg in lightroom publishing services you can add metadata and customised presets).

Searching a photo database …

earching for a specific photo is not such an easy task on a computer right at the moment. The way we recognise things visually is still a challenge for any single computer application to match. Certainly we know a lot about how our brain & visual cortex deconstruct the light that reaches our eyes and how it recognizes shapes and edges and builds a tonal representation, and recognizes colour and saturation. Individually various pattern recognition systems can reproduces one of these tasks at a time today. One on the “holy grails” of computer vision within the artificial intelligence community has been in recognizing objects, potentially those to be manipulated by automated robotic systems, and there are been some successes in very limited spheres of operation. The fact that most smart phones come with a camera has been a great impetus to broader applications of computer image recognition, and already in a few specific fields applications are becoming available, like google’s goggles, which can reasonably accurately recognition a lot of consumer products snapped by smartphones.. But there is still a fair way to go before we have generalized recognition systems that might be really useful in searching through large photos collections and helping organise them are well as any human might. Although I do suspect suspect competition if becoming fierce to solve the more general recognition,  some interesting current development are everpix,, IQ Engines soon to be included in Flickr and here is a brief description of How Google's Image Recognition Works, in Google Drive & Google+. Yet right now

Ok back to the more pragmatic, what is available today.  Finding photos usually relised on a text based search of keywords associated with the photo. I know a lot will say just use lightroom, it manages your collections. I do use lightroom, but it is not really so good at search. It does have keywording and relies on the user to adequately keyword and tag all the photos. The main tool is library/find, which basically gives you a totally text based filter that can browes a subset your catalogue by keywords.attritubes, ratings and EXIF metedata. This all requires a lot of upkeep to keep this useful and there are some good tools to simplify updating the metadata but it remains a largely manual process. A lot of folk overlook the power of sort (again it has filter style dialogue) to help locate and organise (for example it it easy to sort by landscape or portrait orientation or by aspect a ratio if you know you used a square crop for example. the real power is hidden in smart collections, which are dynamically updated  as you classify your collections. LR 4 and above does have a map feature, and it will display any photos with already embedded GPS coordinates or let you drag photos from the filmstrip onto a specific map location, the search capability is just my map window and limited filters such as tagged and untagged photos.

One little package that surprised me with its ability to search was XnView, (which I have as my image scan, sort and keyword tool for my USB darkroom portable apps key (ie a backup when I’m travelling and don't have access to my own computer and picasa). I haven’t written much about this package but it is a gem, Basically it is a thumbnailer, It just creates a database of thumbnails and loads that with all the information it can glean from the photos into its database. It does have a few different ways to look at and organize the thumbnails, Firstly somewhat like the Lightroom filters, it has a single but just about as comprehensive Tools/Search…feature. It also uses a special type of keywording that it called categories which can be set up like a tree structures/classification system (and it comes with some good basic starting categories. The View tabs also have various viewing filters and sorting options which can be useful in sub-setting what you see.

I must admit picasa is my go to application when I am searching through my substantial photo archive. For a few reasons, Firstly it is tolerably fast (lightroom isn’t), second it has face recognition, thirdly because I use its default file naming as I load photos my collection is nicely organised in chronological order. Whilst I seldom use its places features to tag location i do use this feature to view and find photos from a given location. Its search box is in the classic google style although it does just have a text search functionality looking through filenames. folders and album names, keywords(tags), captions place names, and EXIF metadata. It is not really as comprehensive as the previous two but it simple and fast and the screen view changes according to what is available. The filters are there too but are different button/icons that further control what is displayed. While there are no “smart” collections,  people can be automatically grouped by the facial recognition, starred photos are continually updated to a special album, as are recent uploaded photos,

So my conclusion is while the future promises a lot, the present delivers just a enough.(if you are happy to maintain your own keywords)

This post is part one of my rethinking the longer term focus.