Logo: University of Southern California

How Often Are Women Seen/Heard on Screen?

New Software Co-Developed By SAIL Lab Will Tell You
USC Viterbi Staff
September 26, 2016 —


Photo credit: Istock


While it is has been said that there are fewer roles for actresses after a certain age, did you know that even when they are in movies, women are less likely to be heard? That men have twice the amount of screen time? A new software tool co-developed by Viterbi researchers can now automatically assess on-screen gender disparities. The software, developed by the Geena Davis Institute, Google.org, Viterbi's Shri Narayanan and a team of researchers at the Ming Hsieh Department of Electrical Engineering’s Signal Analysis and Interpretation Laboratory (SAIL) for the Geena Davis Institute at Mount Saint Mary’s, can now quantify screen time and speaking time by gender. Funded by Google.org and incorporating Google’s machine learning technology and USC’s audio-visual processing technologies, the Geena Davis Inclusion Quotient (GD-IQ) is the only tool in existence with the ability to measure screen and speaking time through the use of automation. The data tool was meticulously developed over a period of two years in order to address the imbalance of the representation of specific demographics and stereotypes in media.

Existing research on gender, race, and other representations in media almost exclusively employ manually scored content analysis.

“Advances in integrated machine processing of visual and audio content offers immense objective insights of visual and audio content that goes beyond those that are discernable by human ‘looking and listening.’ This opens up new possibilities from a content design to predicting impact of content,” said Shri Narayanan, the founding director of USC’s Signal Analysis and Interpretation Laboratory.

The tool can also calculate content detail with a level of accuracy that eludes human coders. The expectation is that with the speed of a tool like GD-IQ, content creators in film, television, advertising, publishing, digital and more will be able to identify and recognize the issues contributing to the imbalance and correct the course.

The Geena Davis Institute commissioned the tool to encourage inclusivity in media. “Media that is more representative of our society not only fosters a more inclusive industry, but by increasing the number and diversity of female leaders and role models on screen, content creators are affecting the ambitions and career aspirations of young girls and young women everywhere. If she can see it, she can be it,” said Geena Davis.

Using GD-IQ, the Geena Davis Institute generated a new report. GD-IQ’s first study analyzed gender, screen and speaking time from the top 200 grossing (non-animated) films of 2014 and 2015 as listed by Variety. The tool accurately measured on-screen time and speaking time by dividing each movie into a series of shots and detecting the gender of the person on camera by applying an automatic speech detection program that classifies the speaker as female or male. It also captured the total screen time by gender for each film.

The GD-IQ reveals that even when female characters are included, male characters receive significantly more screen time and more speaking time. Our key findings for 2015 show that:

• Male characters receive about two times the amount of screen time as female characters (28.5% compared to 16.0%)
• Male characters speak two times as often as female characters (28.4% compared to 15.4%)

The data indicates that solving gender inequity in in film is far more complex than simply adding more female characters. When they are present, female characters in film are seen and heard far less often than male characters.

Additionally, the study also examined the profitability of female-led films, busting the myth that female-led features don’t perform well at the box office. We find that:
• Films with female leads made 15.8% more on average than films with male leads ($89,941,176 compared to $75,738,095)
• Films featuring male and female co-leads earned 23.5% more on average that films with solo male or solo female leads ($108,317,073 compared to an average of $82,839,635)

Additional findings and detailed methodology for the report are available at http://seejane.org/research-informs-empowers/data/

“Our research shows that when it comes to encouraging women and underrepresented minorities to pursue careers in computer science and other technical fields, positive role models matter, which makes the work of the media industry so powerful,” said Julie Ann Crommett, Google Entertainment Industry Educator in Chief. “

We’re thrilled to partner with the Geena Davis Institute and support this tool to shed light on this important issue and inspire more creators to showcase positive portrayals of women and underrepresented minorities in tech. Thus far, we've worked with many media partners including Disney Jr. to inspire new narratives by sharing the experiences and expertise of our very own Googlers with the writers and creators of various content pieces. Hopefully, this tool is helpful to them and others as they continue to shape new and inspiring narratives."

Further developed automated methods used to analyze individual-level character attributes such as race and age, representations of animated characters, and the composition of background scenes will be demonstrated on on Oct. 18th in Los Angeles.