PhotoRoom, Google Cloud partner to democratize generative AI photo editing

AI-powered photo editing app PhotoRoom has partnered with Google Cloud to use the capabilities of Google Cloud’s A3 instances, powered by Nvidia GPUs, and its expertise in scaling large AI models. The partnership aims to deliver high-quality images to a broader array of businesses through generative AI, and PhotoRoom anticipates that Google Cloud’s support will significantly speed its content delivery.

PhotoRoom claims that its platform’s newly added generative AI capabilities will decrease the time needed to produce photography content, reducing it to under an hour while ensuring optimal accuracy and quality.

The company stated that small businesses (SMBs) and entrepreneurs will greatly profit from the platform’s improved speed and scalability, reducing the time and costs involved in creating and editing commercial photography.

>>Follow VentureBeat’s ongoing generative AI coverage<<

“By leveraging generative AI, we aim to benefit small businesses by enabling them to generate high-quality product photos quickly and affordably,” Matthieu Rouif, cofounder and CEO of PhotoRoom, told VentureBeat. “Our AI model allows users to create entire scenes around a product, all based on a single smartphone photo. This process reduces the need for studio photography, which would cost thousands of dollars to small businesses.”

The partnership builds upon PhotoRoom’s recent launch of its Instant Backgrounds and Instant Shadows features, which use AI to enhance product shots. With these advancements, the company believes it can democratize the use of generative AI, empowering small and medium-sized businesses with cutting-edge technologies.

Using generative AI to enhance photo editing speed

PhotoRoom’s Rouif said that the company incorporated a trained diffusion AI model into its platform to generate images specifically for ecommerce and product photography. The integration of Google’s computing power and memory capacity now enables the application to achieve swifter, more precise and more scalable image creation.

“For small businesses, time is money, and PhotoRoom is 10 times faster than other generative solutions: generating AI images in one second, compared to 15 seconds on average for Midjourney and DALL-E,” Rouif told VentureBeat. “The efficient scaling offered by Google Cloud’s infrastructure enables us to handle an increasing volume of images worldwide without compromising service quality or performance.”

Moreover, PhotoRoom asserts that its Remove Background feature is 30% more accurate than top photo editing alternatives.

“Once the background is removed, small businesses and entrepreneurs require a realistic background to showcase their product — PhotoRoom’s generative AI technology solves this problem by building instant backgrounds and shadows, which help them create content and lifestyle photography,” added Rouif.

The company stated that it currently processes two billion images annually. Collaborating with Google to scale its GPU infrastructure will enable the company to rapidly enhance and broaden its generative AI product offerings.

“This partnership will help us provide a stable and dependable solution to accommodate the increasing adoption of generative AI in the ecommerce industry,” said Rouif. “Google Cloud’s long-term focus on AI and infrastructure investments and open ecosystem aligns well with PhotoRoom’s vision of making high-quality product photography accessible to all.”