Released in The Interline’s AI Report 2024, this executive interview with Vmake AI is one of a fourteen-part series that sees The Interline quiz executives from companies who have either introduced new AI solutions or added meaningful new AI capabilities into their existing platforms
For more on artificial intelligence in fashion, download the full AI Report 2024 completely free of charge and ungated.
What’s your working definition of AI? Does it differ from the public understanding, which is currently dominated by large language models and generative text-to-image models? And how does that definition manifest itself in your solution(s)?
Xuchen Zheng: Our definition of AI extends beyond the popular buzz around large language models and generative visuals. At Vmake AI, we view AI as a transformative tool that mimics human cognitive functions to solve real-world problems. Specifically, we leverage AI to transform visual content creation in e-commerce, focusing on the practical application of generating high-quality, diverse visual models that cater directly to the unique needs of fashion retailers. This goes beyond generating text or images—it’s about creating a dynamic, scalable solution that directly addresses the operational efficiencies our clients desperately need.
Perhaps the biggest roadblock to AI adoption is the difficulty in connecting potential to measurable value in specific process areas. With Vmake AI you’re explicitly targeting eCommerce, so let’s start with a baseline: help our readership understand where delays, waste, and excess cost originate in eCommerce, across content creation and customer acquisition, and how AI can have a tangible positive impact in those areas?
Benben Li: In e-commerce, significant delays and costs are often tied to traditional content creation methods—such as prolonged photo shoots and manual image editing, which can be time-consuming, labor-intensive, and environmentally costly.
Vmake AI’s solution drastically reduces these bottlenecks by streamlining the generation of model images and product visuals. This not only enables rapid online catalog updates and speeds up marketing and customer acquisition efforts, but also reduces the carbon footprint associated with traditional content creation methods, helping businesses scale efficiently and sustainably. For instance, one of our clients has seen a 30% boost in operational output and a 21% increase in 7-day turnover rate. The integration of AI has streamlined their product development process, significantly enhancing product quality by minimizing the need for fine-tuning and simplifying the release of new products.
Diving a little deeper into specifics, tell us how you approached the challenge of building an AI model that can deliver high quality visuals, customisation and brand protection, and the ability to do both on-model content generation and lay-flat images.
Xuchen Zheng: The development of our AI model centered around three key pillars: fidelity, customization, and scalability.
Firstly, we’ve meticulously engineered our AI to produce visuals that adhere to the highest fidelity standards, by employing cutting-edge diffusion models, and training our models on a diverse dataset of high-resolution images. Such an approach enables our models to generate detailed, lifelike images that align precisely with client specifications, which is vital for maintaining brand integrity and building consumer trust for our clients.
Secondly, our deep learning capabilities allow for extensive customization, from accommodating diverse body types to incorporating varied backgrounds, ensuring each image reflects our clients’ unique brand aesthetics accurately.
Lastly, we’ve significantly streamlined the content creation process by supporting both on-model content and lay-flat images. Such dual capability means clients can swiftly switch between displaying their clothing on mannequins and laying them flat for photography, depending on their operational needs. This greatly speeds up the setup time and reduces the manpower typically required for dressing mannequins. This enhancement offers operational flexibility and scales our clients’ ability to produce a higher volume of customized content quickly and efficiently.
Another challenge in content creation and product photography is the economy of scale. Larger organisations, that introduce enough volume and variety of product to conduct their photography in-house, have an inbuilt cost advantage over small-to-medium brands and retailers. How do you see AI helping to even that playing field and allowing smaller merchants to compete more directly with larger ones? And how important do you see this democratising effect of AI being in the future?
Benben Li: You know, there was a time before the invention of electricity when European breweries had to rely on manual brewing methods, which limited their production scale and efficiency. Then electricity was introduced, and the industry underwent a seismic shift. Initially, the largest breweries started building their own power generators, gaining a massive boost in production capacity and cost efficiency. However, as public utilities started providing electricity on demand, even the smallest breweries could scale up production without the prohibitive cost of building and maintaining power infrastructure. This shift democratized brewing, encouraging more focus on crafting better beer rather than worrying about power generation.
Just as electricity revolutionized brewing, AI is transforming the fashion industry. Before AI, retailers, especially smaller ones, struggled with the hefty costs and logistical challenges of traditional content production, from organizing photo shoots to editing images, which favored larger players with more resources. But with the advent of AI, and specifically through Vmake AI’s solutions, the landscape is changing. Our AI solutions act like the utilities of the brewing industry, enabling retailers of all sizes to effortlessly and affordably generate high-quality digital visuals, without needing extensive in-house infrastructure.
This transformational technological advancement allows them to focus on what truly enhances their ‘beer’—their products—by enabling rapid, cost-effective, and high-quality visual content production, much like how utility companies enabled breweries to focus on brewing better beer. As Vmake AI continues to innovate, we are unwavering in our commitment to powering this transformation, ensuring all fashion retailers, regardless of size, can use AI to captivate and engage their customers more effectively than ever before.
For most retailers and brands, growth eventually coincides with global expansion, but this brings with it the challenge of targeting a range of new demographics in a way that’s culturally relevant, inclusive, sensitive and effective. How do you aim to help companies find that balance with AI? What does it look like to build AI systems that are trustworthy and cost-effective, but that also unlock that ability to globalise and localise?
Benben Li: At Vmake AI, we design solutions to cater to global markets with cultural relevance and inclusivity in mind. Our AI offerings are built not only to generate visuals but also to adapt to cultural nuances and demographic specifics, ensuring that the visual content is high quality, culturally attuned, and sensitive. By embedding flexibility and extensive customization capabilities into our AI, we are ensuring our models are free from biases and uphold the values of inclusivity and diversity, which are becoming increasingly essential as brands move into global markets.
At Vmake AI, we proudly offer our clients a diverse range of over 100 AI models. These models encompass a variety of body types, ages, and cultural backgrounds, ensuring inclusivity and representation. We have also introduced a private model service, empowering our clients to create their own unique AI models that truly reflect their brand value and aesthetic. Our approach is empowering our clients to expand globally efficiently without compromising on localization.
What do you see as the near-term future of AI – both within your solution(s) and in general? Do you believe it will be a transformative class of technologies the way people expect?
Benben Li: In the near term, AI is set to radically transform the e-commerce landscape, especially in how online visuals are created and utilized to engage consumers. At Vmake AI, we recognize the intense competitive pressures fashion retailers face, particularly the need for compelling visuals to attract more customers. Vmake AI’s solutions are meticulously designed to address these challenges by powering clients to stand out with high-superior, customizable visuals that significantly enhance online presence and customer engagement.
Xuchen Zheng: As pioneers in applying AI to fashion retail, we are continually advancing our solutions. We have extensively fine-tuned and customized Stable Diffusion and SDXL models to achieve lifelike portrait generation. We also harness the latest advancements from OpenAI to enrich our AI with superior intelligence capabilities in image recognition and analysis. Moreover, our technical prowess extends to proprietary developments such as our semantic segmentation model for clothing, which is highly accurate and designed to enhance the virtual try-on experience. These models allow for dynamic adjustments to clothing items on different body types, providing realistic and appealing presentations.
Benben Li: Looking ahead, we are excited about the rollout of new features and innovations designed to supercharge marketing efforts and customer acquisition. These enhancements will allow our clients to drastically amplify their marketing reach, leveraging the very best of what AI can offer. Our unwavering commitment to deep technological innovation in AI ensures that Vmake AI not only responds to current industry demands but also shapes future trends. We invest heavily in our technology to uphold our position at the forefront of the fashion and e-commerce industries, driving transformative change and delivering unparalleled value to our clients.
