Immediately, Powder, the AI-powered video editing program, is poised to enhance its capabilities by identifying instances of gamers shouting to create even more compelling montages. Powder leverages highlights from gaming streams to transform them into concise video clips. The platform is also in the process of developing speech-to-text functionality, enabling stream creators to search for specific keywords and access transcripts of their entire streams.
For popular games such as Fortnite, Valorant, Apex Legends, Call of Duty, Rocket League, Fall Guys, Elden Ring, and Among Us, Powder has developed over 40 customized AI models tailored to each game, including features like music detection and laughter recognition. Furthermore, the company is preparing to unveil a dedicated model for Counter-Strike2.
All these AI models function in a similar manner. They analyze gaming stream recordings from platforms like Twitch, YouTube, or MP4 files, identifying key moments such as victories, assists, kills, and other noteworthy in-game events. Utilizing this data, Powder generates short montages that creators can share on social media platforms.
Image Credit: Powder
In the near future, Powder will introduce a new AI tool capable of recognizing variations in voice patterns, allowing creators to compile clips of intense moments like shouting during competitive matches, in addition to laughter detection. The anticipated launch for this feature is mid-December.
Barthélémy Kiss, co-founder and CEO of Powder, emphasized the subjective nature of capturing the best gaming moments, extending beyond mere gameplay. He stated, “From uncontrollable laughter to rage quits, even moments with no visible action on-screen, we aim to encapsulate the essence of gaming camaraderie.” The blend of skill-based gameplay and emotionally charged moments makes gaming content creation a unique and special endeavor.
Next month, the platform will introduce speech-to-text capabilities, enabling creators to swiftly search for specific words and retrieve highlight clips while also archiving their streams. Streamers can also input emotional cues, such as “Show me five hilarious clips of viewer reactions,” to facilitate more precise search results, tailored with gaming terminology.
According to Kiss, the ultimate goal for content creators and their support networks, including viewers, agents, and managers, is leveraging AI to analyze and comprehend lengthy video content like Twitch streams.
Moreover, Powder is set to launch an updated version of its “Community Hype” feature, scheduled to roll out the following week. This enhancement builds upon the AI’s ability to detect spikes in chat activity, introduced in September. Kiss explained that this update will recommend videos capturing moments when the community “goes crazy.”
He elaborated, “The second phase of Community Hype recognition aims to provide an additional perspective on what constitutes a stream’s ‘highlight moments.’ Understanding the community’s perspective, as they are intimately familiar with the streamer, is crucial. Communities possess a keen awareness of significant moments during a gaming session or on a particular channel. When the community erupts with excitement and wants to commemorate a moment, Powder AI will suggest saving it as a highlight to share.”
Powder’s survey of over 3,200 streamers revealed that developers spend an average of 53 hours per month or 630 hours per year searching for features and editing videos. Powder claims it can save creators up to 520 hours annually, equivalent to 10 hours per week.
Established in 2018 by Christian Navelot, Yannis Mangematin, and Barthélémy Kiss, the France-based company has raised $22 million in funding thus far.
