InsightPersona

People in media: find, identify and analyze

Which politician was visible how often on which channels – and with which statements? How equally do all genders appear on a particular program? How can I find all the programs featuring a particular person in my archive?  Large media archives offer an enormous treasure trove of data, but it is only with powerful analysis tools that this data can be used and analyzed in an efficient way.

InsightPersona provides insights into the media presence of politicians, celebrities and other individuals or groups on TV, in streams and video archives. Individuals and voices are quickly located and identified through simultaneous facial and speaker analysis. In combination with intelligent speech recognition, spoken words can also be searched for specific content.

With InsightPersona, you can exploit the full potential of your media content – and make optimum use of your research resources. Optimize your reporting, content strategy and research with concrete quotes, meaningful statistics and comprehensible visualizations.

Application areas

© istock.com/miodrag ignjatovic

Create solid reports with little research effort 

Content creators, editors, and journalists can find relevant references, such as quotes and sound bites, on current topics and events within seconds for their reporting. Thousands of hours of video material are automatically searched and analyzed in their database for this purpose.

© istock.com/Jacob Wackerhausen

Optimize your own content strategy with the right insights

Broadcasters, streaming platforms and media companies gain new insights through the comprehensive analysis of their program content and can derive important information for their content strategy as well as for diversity and inclusion.  

© istock.com/microgen

Keep an eye on focus groups and sectors

Agencies for opinion research and companies can create media analyses. Evaluations for scientific studies and reports are facilitated by simple visualizations and statistics.

The features of InsightPersona at a glance

Prepare content

Analyze large volumes of data from TV and radio stations or streaming platforms efficiently: flexibly and comprehensively depending on the application – regardless of whether your media content needs to be recorded in real time or aggregated offline from various libraries and archives.

Simultaneous analysis and interpretation of audiovisual data

The cross-modal analysis of image and audio content enables an improved and more comprehensive understanding of media content. Faces and voices are localized and identified simultaneously. This is particularly helpful in situations where audio and visual data complement each other or enable different perspectives on the data and analysis results.

Audio analysis

Advanced features to analyze audio content, identify speakers, and children's voices, classify perceived gender, convert speech to text, and perform keyword analysis. Additional capabilities enable the evaluation of speech intelligibility and the level of excitement in speech.

Advanced video analysis

Elevate the comprehension of your audiovisual data with AI-based face detection, identification and cutting-edge “perceived gender” classification tools with advanced features and high levels of precision. 

Interactive data dashboard                      

Use interactive dashboards and tools, such as timelines and heatmaps, to visualize and prepare your data, access statistics and identify trends. Transform complex data into intuitive, easy-to-understand visualizations, with data dashboards that are fully customizable.

Perceived Gender Classification

In addition, an evaluation can be carried out based on face and voice, which examines the appearance of people perceived as female and male. Among other things, this function enables detailed analyses of diversity in media content.

The analysis results can be displayed in "heat maps“ that visually represent when and where a person appeared, or was audible, across various TV channels over customer-specified durations.
The “Appearance Details” show when a specific person was visually or audibly present in a program. The combined analysis also indicates when the person was only heard but not displayed in the image, e.g. when he or she speaks in talk shows, but other people of the conversation group or from the audience are faded in.

Learn more

 

Facial Analysis Technologies

Detect faces, recognize a specific person and identify characteristics such as perceived gender or age with our AI-based facial recognition and analysis.

 

Speaker Recognition

Who actually speaks when and where? What language is being spoken? We can answer these questions with our acoustic recognition technology.

 

Better understanding

Analysis and optimization of speech intelligibility