Using film audio, we calculate the latest speaking duration of male and female characters in order to get an objective indicator of sex signal. New formula getting performing this data comes to automated voice interest identification, sounds segmentation, and gender class. Sound Activity Recognition: Motion picture songs usually includes of several non-speech regions, also sound […]
Using film audio, we calculate the latest speaking duration of male and female characters in order to get an objective indicator of sex signal. New formula getting performing this data comes to automated voice interest identification, sounds segmentation, and gender class.
Sound Activity Recognition:
Motion picture songs usually includes 
of several non-speech regions, also sound files, music, and you can silence. The first step is always to lose low-message countries from the musical using voice hobby identification (VAD) and you will hold simply address segments. I put a perennial neural system dependent VAD algorithm followed inside the this new discover-source toolkit OpenSMILE so you’re able to split up speech areas.
SEGMENTATION:
I following crack address markets with the quicker sections to guarantee each section includes speech regarding only 1 presenter. It is did having fun with an algorithm based on Bayes Information Standard (BIC), obtainable in this new KALDI toolkit. 13 dimensional Mel Volume Cepstral Coefficient (MFCC) enjoys can be used for the automated audio speaker segmentation. This step basically decomposes continued speech avenues acquired throughout the VAD step towards the faster markets to ensure zero part includes message out-of one or two various other sound system.
Intercourse Group
The brand new address segment will be classified for the several groups considering in the event it is actually likely spoken by the a female or male character. They do this having acoustic function extraction and feature normalization.
ACOUSTIC Function Removal
We have fun with 13-dimensional MFCC features getting intercourse category as they can be reliably taken from flick tunes, instead of pitch or any other large-peak enjoys where removal is generated unsound by diverse and you can loud nature regarding movie tunes.
Element NORMALIZATION
Element normalization is deemed needed to address the difficulty regarding variability out-of message around the various other movies and you will sound system, and to slow down the effect of audio contained in this new musical station. Cepstral Mean Normalization (CMN) is a standard techniques well-known within the Automated Speech Detection (ASR) or any other message technology programs. This way, the latest cepstral coefficients is linearly switched to obtain the exact same segmental analytics (zero mean).Category of your own speaker because either person depends to your gender-particular Gaussian mix activities (GMMs) of the acoustic enjoys. Such patterns is actually educated for the a gender-annotated subset from general address databases useful developing message development playing with frame-level enjoys each intercourse. The new GMM i use in this program has one hundred combination components that is optimized because of the tuning the latest details when you look at the a held-out research put. For another input segment whose gender label will be predicted, the latest likelihoods of your phase owned by a female or male category try determined predicated on it pre-instructed model. The course that have large likelihood belongs to the newest section once the the estimated intercourse prediction. The total speaking time by the sex is then calculated adding together the fresh new periods per utterance classified since Men/Girls. This gives united states the male and you can people talking time in a good film.
step 3. Objectification so much more broadly mode dealing with one as a commodity or an object in place of regard to their identification otherwise dignity. Panning means rotating a digital camera towards the its straight otherwise horizontal axis. In this situation, they identifies moving from just one element of a human anatomy so you’re able to various other. Slow-motion are often used to complement individuals regions of the fresh photographs into the a screen. For it form of level, list instances when slow-motion can be used so you’re able to enhance a good character’s physical function inside the a sexual way, instance, jiggling boobs. Spoken intimate objectification can come in several forms, together with pet getting in touch with and statements a character can make in the another character’s physicality in order to a 3rd party.
4. Pick Levant, Roentgen. F., Hirsch, L. S., Celentano, Age., & Cozza, T. Yards. (1992). ”The male Character: An investigation of contemporary Norms.” Journal off Psychological state Counseling and you will Mother, 14(3), 325-37. Find along with Meters. C., & Moradi, B. (2011). “An Abbreviated Device to possess Determining Compliance in order to Masculine Norms: Psychometric Characteristics of Conformity so you can Male Norms List,” Therapy of men & Maleness, 12(4), 339.
veja mais