I am working on facial emotion recognition task, for that I need either dynamic facial images or morphed one as an experimental stimuli (As static images I have already used). For that purpose, please do suggest some concerned database.
Morning Kushi,
I have read some work on:
https://www.ri.cmu.edu/research_project_detail.html?project_id=418&menu_id=261
This should assist, but on your requirement, it might result in the need for some time to be spent based on your IDE (if it supports)/Project requirement. Another alternative for you is to simply use a MySQL blob - convert each of these files (whether vector/object in any format, if they work fine in your container) into binary - encrypt them and store as blobs when you need persistent storage.
For retrieval, do the same process in reverse - do not it might not be as efficient but you may have more control over these objects and how they may be viewed. You might need to manage the file at every stage though.
Hope this helps.
Hi,
you can use Facial Action Coding System (FACS). In the manual they use a set of images and videos to show how the muscle movement reflects the emotions. There was also a program to train emotion recognition, but I think it was a not academic software.
Here is also a webpage which allows you to manipulate muscles/emotions. Could be interesting for you (http://www.artnatomia.net/uk/artnatomy2014.html). Go to "Application" - "Naturalistic model" - "Action".
Another option is to cut off some scenes from videos and rate them by professionals before the test.
i think you can find some information in you-tube:
https://www.google.com/search?q=dynamic+facial+images+or+morphed+images+for+facial+emotion+recognition+task&ie=utf-8&oe=utf-8#q=facial+emotion+recognition+task+youtube
If you're looking for dynamic (video) as opposed to static (pictures) there's one that will definitely work within the basic/discrete emotions theory: http://www.psych.qub.ac.uk/BINED/ Best I've seen so far. Cool methodology, excellent metadata, responsive team, high-quality material.
I created morphs of facial expression along a number of scales as part of my PhD covering the development of an instrument to measure mood nonverbally. The scales and their images are available at http://www.dvams.com.
If you use any of my images as part of your research, please include the copyright notice: "Copyright © 2013 Paul Barrows".
Hi Kushi, you may find the RAVDESS helpful in your paradigm. It's a validated multimodal database of emotional speech and song. It contain 7356 recordings in English, with 8 emotions: calm, happy, sad, angry, fearful, surprise, disgust, and neutral, each at two emotional intensities.
Article The Ryerson Audio-Visual Database of Emotional Speech and So...
Downloaded for free - https://zenodo.org/record/1188976