Suggestion: Photo Quality Rank

[Cullen Hanks]

I was looking through the eBird media guide (http://help.ebird.org/customer/portal/articles/973966-adding-photos-videos-and-recordings-to-checklists), and I like the way they rank the photos separately from the observation quality or value.  This might have value for the iNaturalist community. 

For example, if you wanted to find the best quality photos of an organism to compare variations within the species, or just to solicit the use of a good photo for a presentation, it would be nice to be able to search for observations of a species with high quality photos. 

This is not a big need for us at TPWD, but I thought it was worth putting the idea out there.

Best,

Cullen 

[Ken-ichi]

We were hoping faves would solve this problem, i.e. that people would
fave photos that were aesthetically pleasing, but people seem to fave
for a number of different reasons (interesting subject, user-specific
categorization, etc. You can generally use fave sorting to get some
decent pics toward the top, thought, e.g.

http://www.inaturalist.org/observations?taxon_name=Coleoptera&order_by=votes&order=desc

> --
> You received this message because you are subscribed to the Google Groups
> "iNaturalist" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to inaturalist...@googlegroups.com.
> To post to this group, send email to inatu...@googlegroups.com.
> Visit this group at http://groups.google.com/group/inaturalist.
> For more options, visit https://groups.google.com/d/optout.

[James Bailey]

If I faved for every nice picture I saw, I'd have hundreds of pages of favourites. It has to be a truly exceptional photograph for me to fave on that alone.

Though a "star rating" system for photos sounds nice (as proven by the high number of sites that have this system), I'm not sure it would really achieve much.

[Cullen Hanks]

It would allow people to better sift through the clutter to view quality images of a particular taxa in a particular region, and that has value.  Here is a screenshot of the ebird media upload that not only encourages a rank of the photo, but also comments on a particular photo.  This would have value when you have multiple photos that highlight a different identification characteristics or behaviors.  Granted, this could be described in the observation main field, but it is worth thinking about.

To me, the fav function does not have much value because it is undefined and people can fav an observation for many different reasons. 

 -Cullen

[SummitMetroParks-NaturalResources]

I use faves, both as intended, as well as to "flag" records that I want to get back to later.  (I do really like the star idea a lot to rank photos on how well they present the subject, but that is not my reason for commenting)  My preference would be to use faves, as I believe they are primarily being used, to represent someones favorite observations.  It would seem that this could potentially replace the "favorite taxa" section, especially if the faves were grouped into like categories.

My comments are really leading to the point that I feel we should have some sort of personal flagging system, like faves, but just for personal use to get back to records that need more research, or that we want for reference, etc.  I am pretty sure many (including me) are presently using the faves tool to accomplish this very thing.  But such a practice just takes away from its original intent.  Does anyone else feel a need to mark/"flag" certain observations for future reference?

[Colleen Boye]

This seems a little catty to me. We wouldn't want to discourage a beginner with a phone camera because all her observations get 1-star reviews. Concern that the quality of one's observations isn't good enough and "oh, nobody would want to see those silly pictures I took" is already a reason people don't participate.

Possible solution: Instead of a rating system, just have a way to vote for an observation as a "top observation" or "exceptional observation" of a particular species. It can have the same desired effect--moving really good photos to the top of the results--but it would frame the whole thing in positive terms and be more about rewarding the best observations instead of punishing the worst ones.

[Carrie Seltzer]

I think Colleen makes a good point about not wanting to discourage people by rating all photos. I think "faves" could still serve this function if there were a bookmarking feature that allowed people to separate "awesome/exceptional/beautiful" from "something I want to revisit". 

For what it's worth, photos with the most faves are highlighted in the BioBlitz slideshows that are automatically generated for NPS BioBlitzes, so favoriting nice photos from U.S. National Parks is especially useful. It takes a couple of minutes to get to the faves slide (it changes automatically and you can't control the timing of the slides), but you can check it out for the servicewide project here: http://www.inaturalist.org/stats/nps_bioblitz

For those of you with little patience, here's a screen shot of what it looks like right now: 

If you want to see what else already has some faves in from NPS and add some more, go here: http://www.inaturalist.org/observations?place_id=any&popular&project_id=6810&subview=grid&verifiable=any

Carrie

[Kent McFarland]

On the iNaturalist Vermont project we use Favs each month to elect the photo-observation of the month. 

Here's a few from the last months as examples:

http://vtecostudies.org/blog/june-2016-inaturalist-vermont-photo-observation-of-the-month/

http://vtecostudies.org/blog/may-2016-inaturalist-vermont-photo-observation-of-the-month/

http://vtecostudies.org/blog/april-2016-inaturalist-vermont-photo-observation-of-the-month/

Kent

____________________________

Kent McFarland
Vermont Center for Ecostudies
PO Box 420 | Norwich, Vermont 05055
802.649.1431 x2

--

You received this message because you are subscribed to the Google Groups "iNaturalist" group.
To unsubscribe from this group and stop receiving emails from it, send an email to inaturalist...@googlegroups.com.
To post to this group, send email to inatu...@googlegroups.com.

Visit this group at https://groups.google.com/group/inaturalist.

[James Bailey]

I'd rather have a system where you give a star to a photo similar to how you give a fav or "like" on some other sites. That way it is less of an issue in regards to the scenario of "oh this photo is amazing it has 5 stars, but hm well your one isn't good so its only 1 star".

But this star would be meant only for good quality photos.

[Lindsey Wise]

I'd say 80% of my faves are things I want to come back to or are of interest not for the photo. So having a bookmarked list separate from a "great photo" list would work well for me.

-Lindsey

[Colleen Boye]

It seems like we're all on the same page here: Something similar to likes/favs, but that clearly applies only to photo quality and not to all the other reasons you might fav an observation. (Also note that this should apply to photos, not to observations, since the latter might have multiple photos of varying quality.)

[Julien Renoult]

Hi, 

I come back to this "old" thread here because I wanted to post a message proposing exactly what Cullen proposed.

Now that the number of pictures on iNat is really huge, I think it becomes important to separate the quality of the picture from the quality of the observation. A poor-quality, distant picture is extremely valuable if it allows confirming identification. But a good-quality picture has additional advantages, e.g. it can help to taxonomic works by documenting interspecific variation. This would be great to be able to identify quickly those good-quality pictures, using a filter. I also suggest that good-quality pictures come first when one do not sort by observation date (for example in the photo gallery of taxon pages).

I think a 3-star scale with a default value of one star would be good. Less "violent" than a 5-star scale ranging from 0 to 5. I think that, compared to a simple "like" option, a real scale, even if it has only 3 different values, make it more obvious for users that what is expected here is the RATE the quality of the picture. A simple "like" may have the same caveat as the "favorite" option already implemented: it is not obvious on which basis we should like it or not.

My best

Julien  

[Mark McGrouther]

Your 3 star rating has a lot of merit.

I like it.

Mark

[Cullen Hanks]

Julien, thanks for revisiting this thread!  The fave functionality is so broad that it does not mean anything specifically.  You can fave an observation for so many valid reasons. Furthermore the "most exceptional" observations are often geographic outliers, and not necessarily the best representations of an organism.

I agree with others that suggest there should be a ranking option associated with the media, not the observation.  Keep the fave functionality for observations if that is of value, but add a media quality ranking.  This would allow you to identify good quality media representing a species.  Anyway, some observations can have high and low quality photos.

As an example, you might have a normal observation of a common species with two photos.  One photo might be of poor quality, but it shows an interesting behavior.  The second photo could be exceptional, and an excellent representation of the species.  Depending on the behavior, you or others may or may not want to "fav" the observation, but the poor quality photo would still get 1 star, and the high quality photo would get 3 stars.

I like the simplicity of a 3 star ranking, but might prefer 4 stars.  For the sake of thought, here is how I might break it down:

1 star: Poor quality, organism not visible or not a good representation of the species; added for identification, behavior, or other reason.  Organism is out of focus, does not make up much of the frame, blurry, partially blocked from view, and/or is in bad light.  1 star would also apply to any photo of habitat or otherwise does not explicitly feature the organism identified in the observation.  (For example, an observation may have a diagnostic sound recording along with a photo of the habitat.)

2 stars:Average quality: Organism is decently in focus, takes up a good percentage of the frame, has enough light to see detail, and it is easy to differentiate the organism from the background and other individuals.  (This might be the default ranking?)

3 Stars: Good Quality: Like average quality but the focus is crisp and it is in good light.

4 Stars: Exceptional: The best photos in quality and aesthetic.  In addition to being technically high quality, these photos are good representations of the species in regards to behavior, identification, etc...  These photos will be featured first when a person wants to see representative photos of a species. 

Regarding the impact on new users:

Maybe don't lead with the ranking and make it optional.  It should be value added, and not part of the primary user feedback interface.  That said, I wish I could rank the quality of my own photos, especially my poor quality photos.  Maybe only allow others to give 2 or more stars, and only you can give your own photo 1 star.  Maybe the rank of photos should be private, and not part of the feedback environment for observers (I don't like this solution). I think the best solution is to not highlight the photo quality rank except during photo upload.  This is a legitimate issue, but I think there is plenty of room for balancing the user experience of new users and our ability to identify the best representations of an organism.

Best,

Cullen

  

--

You received this message because you are subscribed to the Google Groups "iNaturalist" group.

To unsubscribe from this group and stop receiving emails from it, send an email to inaturalist+unsubscribe@googlegroups.com.

[jesse rorabaugh]

I still have concern that a star system would serve to demotivate people as they get lots of low star numbers. There are a lot of submissions which are poor quality photos but still are identifiable. We certainly don't want to discourage those from being submitted (although poor unidentifiable photos are another matter). Therefore I really do think that a strictly positive system would end up working better. Maybe something analogous to the rating system for yelp comments where you can rate a comment as useful, funny, or cool.

[Charlie Hohn]

I would use this but i agree that it should not be phrased in a way that discourages 'bad' photos. I suppose i am biased because i take a lot of less than beautiful photos because i care more about documenting presence of things than getting nice photos - all i care is that it can get RG. However i myself am not bothered by those highway white pine observations and such getting just one star. Maybe just bury it a little so that people only find it when they are really looking. By then they should understand the purpose.  Like Jesse said an alternative would be just to be able to tag photos as particularly good/diagnostic. Maybe include a way to note which features are present (leaves of sugar maple showing opposite placement, bird is male and in breeding plumage, etc) but maybe that is too complex and better addressed in fields.

In the short term one could make a 'photo quality' field to mark good quality photos of course but since fields go per observation and not per photo it could be confusing when there is more than one photo

[James Bailey]

I only see an obvious need for two levels of photos. Level 1: the photo is clear and useable. Level 2: the organism is not visible or is blurry or otherwise cannot be identified because of bad quality.

For amazing photos, we have the favourite system sort of working for that. Fundamentally, I feel those two levels are the most beneficial for the site.

[Cullen Hanks]

James et al.

If you want to waste a bunch of time, just sort by Faves, and indeed you see some amazing images and observations.  Really awesome and inspiring to look through these observations.

http://www.inaturalist.org/observations?order_by=votes&subview=grid

iNaturalist is creating an incredible repository of images.  It would be great to pull up the high quality images for learning and comparison purposes.  Faves is better than nothing, but it is not specific to image quality or to the actual image.  I worry that the lack of quality evaluation will limit the utility of these images for learning and research applications in the long run.

This becomes more apparent when you use Faves to pull up the best images for a particular species.  With some species, like predators, faves can more likely to be used for observations that document a cool behavior rather than the organism.  In these cases faves may be a really poor tool for selecting the best representations of the species.  Here are three examples picked at random:

http://www.inaturalist.org/observations?order_by=votes&taxon_id=47075

http://www.inaturalist.org/observations?order_by=votes&taxon_id=9083

http://www.inaturalist.org/observations?order_by=votes&taxon_id=42051

http://www.inaturalist.org/observations?order_by=votes&taxon_id=4647

If there is an aversion to ranking images, another approach could be to have a gold star ranking.  Something to indicate that a particular image is a good quality representation of the species. 1=decent 2=good 3=excellent

It could be on the photo details page, not on the observation, ex: http://www.inaturalist.org/photos/5851005

The good news is that unlike some attributes, this functionality and attribution can always be added later.  

Best,

Cullen