Emotional Dataset Creation

Here you can find details about design of this dataset and the results we had and used for the paper 'Combining Semantics and Acoustics Features for Valence and Arousal Recognition in Speech'. If you want to participate in the experiment please read below and notify me at 'seka AT imm.dtu.dk'.

You will be provided with the instructions in the beginning of each experiment. However, in case you may need it during the experiments, you can download it from this link : Instructions.
Please use a headphone during the experiments and try to be in an undisturbing environment.
You will see below the window with a start button on it. Fill in the personal information part with your name, occupation, gender, age and english fluency information. After you enter all the information needed, the 'Start' button will be highlighted and please press it to start. (Note that you need to fill in all the information needed for the start button to be enabled and it may take a few seconds to around a minute to load depending on your net connection speed)
There is also a 'continue' button. If for a reason you had to close the webpage, or internet connection problem etc, and you couldn't finish the experiments, you can continue from where you left by pressing this button. Please note that you have to enter the same name you entered before.
There are 3 experiments. Please do all the experiments in the order EXP 1, EXP 2 and EXP 3. Once you finish one experiment, the next one will be highlighted. After finishing the experiments, you are kindly required to fill in a very short questionnaire. The "Questionnaire" button will be enabled after you finish EXP 3.
Please try not to give long breaks between clips or experiments. You can have short breaks though. We take the track of your rating time to check the validity of the results (long breaks would ruin the validity of data).