ACQUINE - Aesthetic Quality Inference Engine
Dr. Datta, Prof. Li, and Prof. Wang ACQUINE (Aesthetic Quality Inference Engine) is a machine-learning based online system of computer-based prediction of aesthetic quality for color natural photographic pictures. Because of the very limited computational resources and the real-time processing requirement, much complexity reduction in this implementation is necessary. The system demonstrated here is therefore mainly for assessing the potential of such research by the research community and the general public. In our opinion, even though Acquine has much room for improvements, this is an important step in the intersection of computer science research and the arts because it shows that computers can learn about and exhibit some "emotional responses" to visual stimulus like humans do.

Acquine has been developed at Penn State (home of the Nittany Lions) since about 2005. Dr. Ritendra Datta (now with the Google engineering office in Pittsburgh) was the main developer, working with his thesis advisers Prof. James Z. Wang and Prof. Jia Li. The system was placed online for public use in April 2009. Dr. Dhiraj Joshi (now with the Kodak Research Labs) contributed to an earlier prototype when he was with the research group. Since 2010, the research group members are actively contributing to the the future improvements. For instance, honors undergraduate student Razvan Orendovici is working on the user interaction, integration of other technologies, among other things.

Acquine is work-in-progress and hence it undergoes algorithmic changes from time to time, in an effort to improve performance. The work is Patent Pending. If you are interested in licensing this technology from Penn State, please contact James Wang.

Because of the limitations on the sources from which Acquine was able to gain some understanding about aesthetics, the opinions expressed by Acquine can be biased by the group of people associated with the sources. Whereas Acquine is possibly less biased than individual people at the time of photo assessment, there is no absolute unbiased opinions on aesthetics.

What is Acquine Designed to Handle?
Acquine is designed mainly to assess the aesthetic quality of color natural professional photographs. You can choose to use it to assess other types of photographs or images, such as those listed below. Any meaningful use of Acquine is all right.

Acquine is not designed for assessing certain image types, including the following:

  1. computer graphics,
  2. artificially-produced diagrams,
  3. figures in publications,
  4. paintings,
  5. composite pictures,
  6. casual family photos,
  7. screenshots,
  8. out-of-focus shots,
  9. advertisement images,
  10. photos of industrial products,
  11. cartoons,
  12. political photos,
  13. news photos,
  14. etc.
At the moment, Acquine cannot understand the great complexity of our human society and should not be used for assessing photos with a lot of cultural meanings. Please do not upload objectionable photos or any photo containing private information that you do not wish to share with others. We remove uploaded photos inappropriate for this site. Full color photos are more suitable than black-and-white photos. The system is not designed to assess very low resolution images. Our recommended resolution is for photo images with 600x600 or more pixels. Acquine is not designed to rate the look or the attractiveness of a person or a product.

As for most computer-based systems, it is possible to find special cases where the system is clearly not functioning as intended. For instance, one may find that a very poor quality photo gets a good score, or an award-winning photo gets a low score. A rule of thumb is that if the aesthetic quality of a photo is obvious to most people, it may not be worthwhile to seek Acquine's opinion on it because Acquine may choose to assign funny scores in such cases. Please be serious if you would like Acquine to help you.

How to Use Acquine to Score Photographs:
There are two simple ways in which Acquine can be seen in action: Why are Some Reasonable Photos Rated so Poorly?
Whereas it is not rocket science, automatic assessing photo aesthetics is a new scientific research area and is not simple. The research is still on-going. We do not expect our friends to always agree with us on visual aesthetics. We should therefore not expect Acquine to always agree with us on visual aesthetics. We are essentially trying to "teach" Acquine about those human emotional reactions to visual stimulus. Aesthetics is just an important first step in this direction. It can take time for the Acquine to learn about our world. So please be patient. If your photos are not rated as highly as you would expect, it may not always mean that your photos are not of high aesthetic quality. You are encouraged to upload such photos to sites like photo.net to get human assessments and critiques.

When humans assess the aesthetic quality of photos, score inflation is commonly seen. This is probably because many of us want to be more encouraging to others. However, Acquine tries to give you a score that is not inflated. That is, the score actually goes from zero to 100, with 50 being the score of a typical average-quality professional photograph. In another word, a photo with a score of 50 is still considered by Acquine a good picture. See the score distributions of all uploaded photos. Acquine is tough, even to photos taken by the developers!

Some may have noticed that photos with frames can get somewhat different Acquine scores from those of the original photos. Whereas we would discourage adding frames to photos submitted to Acquine just for the sake of getting a better reading (in fact, many photos without frames are also scored highly), it is known to the art community that the presentation of an art piece can be important. French painter Edouard Manet said, "Without the proper frame, the artist loses one hundred per cent." Another French artist, Edgar Degas, put it in a positive way, "The frame is the reward for the artist." So if you are to present your photos to other people, consider whether adding the right frames might help.

Services and Acceptable Use Policy:
The services provided on this site are for human interactive use only. Due to limited computing resources, any automated use is strictly prohibited unless prior approval is obtained. We will try to keep the services on this Website free and free of advertisement.

Please do not upload materials that can be offensive to other users of the services. Do not purposely check off wrong tags after an image is uploaded. We reserve the rights to terminate your use or block access to your pictures if such activities are detected. Please try to upload pictures of good taste.



How to Paste in a URL of an Image on the Internet:

If you are a user of Flickr, you can get the URL of most Flickr images. First use the View Image function when the menu pops up, then copy the URL of the image from the browser's URL field.
You can try to drag and drop a file into the file upload field. Many browsers support this.

Main Scientific Publications Include
Results on photo.net pictures: photo.net [ random, ranked ]

Contact: If you would like to report abuse on this site, request deletion of images, report our work in the news, or have other questions, please feel free to contact Prof. James Wang (, +1-814-865-7889). A Penn State news release (05/2009) is available here. To delete your photos, please send us the IP address of the machine you used to upload, or the URL of one of your photos, as stored on our site.

Sponsorship: The National Science Foundation has funded past work related to this project. We are currently seeking funding to significantly improve this technology and make it more useful. If you would like to sponsor or partner with us, please contact James Wang. We are in serious need of hardware support to make this site useful to the public. If you work for one of the hardware companies and would like to sponsor some storage, server, and networking equipment, we will put a logo of your company on this site.


© alipr.com Acquine