Is that drawing the real deal or not? Artrendex’s AI can likely tell. It has called the bluff on paintings of questionable provenance and confirmed that a master was behind a previously undocumented work.

Developed and refined over the last five years, Artrendex’s patented technology can determine whether a painting, drawing, or print exhibits the micro-characteristics that are telltale signs of a certain artist’s hand—or if they are the work of a copycat. By minutely examining lines, shading, and other qualities, Artrendex has authenticated works by Picasso, Cezanne, Basquiat, and many others for galleries, museums, and other institutions. 

“AI can be a life-changing tool for curators, valuation experts, and art historians,” says Ahmed Elgammal, founder of Artrendex and head of the Art & AI Laboratory at Rutgers University. “It’s excellent at detecting patterns in complex data like hatching, brush strokes, and other telltale signs of an artist’s method. With savvy training and the right data sets, AI models like ours can offer a whole new horizon for art authentication.” 

Artrendex AI authentication uses deep learning techniques, but is vastly different from off-the-shelf products that recognize faces or objects in images. “We cannot just use AI models that recognize cats from dogs and train them to recognize Van Gogh! Such models will be easily fooled by forgeries. We need to go to the stroke level and learn the spontaneous characteristics of the artist’s strokes,” says Elgammal.

Art authentication involves a great deal of expertise, and AI builds on that deep knowledge of art history and materials. AI’s ability to spot patterns with a precision difficult for humans means it can “see” what people may overlook. Training a model that can authenticate accurately and detect copies requires finding patterns at the stroke/line level and connecting them to artists’ methods and era, a key step many AI authentication companies miss. 

“You need to make sure you’re not including any potential forgeries or copies in your data set,” Elgammal explains, “and you need to be honest about your confidence level, which should be extremely high. Ours ranges from 90-97% in determining whether something is by a particular artist or not.” 

To test the Artrendex model, Elgammal and his team compared the works in two exhibits of Basquiat’s work, one in New York and the other in Miami. The New York paintings all appeared to be genuine, while the AI flagged the Miami paintings as fakes. These results were later confirmed independently as part of an extensive investigation by federal authorities. 

This interplay of expertise and AI precision allowed Artrendex to identify a painting an art lover spotted at an architect and art collector’s estate sale in Arizona in early 2000, which included many other works of note. As the art lover looked over the unidentified painting, he saw what he thought was the mark of a master. Without provenance, he had no way of definitively identifying the artist, though he suspected it might be Pablo Picasso. Consulting with Spanish Picasso experts, including Dr. Sergio Ruiz-Moreno from Polytechnic University of Catalunya, the work’s owner allowed the painting to be examined closely and tested for physical attributes like pigment age. All elements indicated that the painting was indeed a Picasso.

To further confirm this, the owner turned to Dr. Elgammal and Artrendex’s AI tools, Artrendex examined the work’s characteristics from yet another angle. The research aimed to measure and analyze the statistical attributes of the distinct pen and brushstrokes in the artwork under investigation. Agreeing with the conclusion of the human experts, AI helped confirm the owner’s instinct that the work was indeed a Picasso. (Artrendex provides only opinion based on scientific evaluation. The final decision on authentication is due to the Picasso estate.)

Ideally, an AI model like Artrendex could be used by online and brick-and-mortar art galleries to help authenticate sellers’ offerings or could guide a museum in a purchase decision. “When deciding whether to purchase a work, buyers need every tool they can get to inform their choice. The art world has resisted this, but the time has come for a change,” explains Elgammal. “Artrendex is one of those tools, a highly reliable, research-based way to double check authenticity and gain insights into the fine details of an artwork.”

About Artrendex
Artrendex is a New York-based technology company that specializes in using artificial intelligence (AI) to enhance the art industry. The company was founded in 2018 with the aim of using machine learning and data analysis to revolutionize the way people discover and engage with art. Artrendex's flagship product is a visual search engine called ArtPI, which allows users to search for art by visual similarity rather than just text-based keywords. ArtPI uses a combination of computer vision and machine learning algorithms to analyze the visual characteristics of an artwork, such as color, texture, and shape, and then recommend similar works of art to the user. In addition to its visual search engine, Artrendex has also developed other AI-powered tools for the art industry featured on its AI creativity platform for visual artists Playform. Overall, Artrendex's mission is to make art more accessible and engaging for everyone, and to use the power of AI to bring people closer to the art they love.

ArtPI (artpi.co) is a new interface or API driven by artificial intelligence that’s poised to transform the way art gets discovered, displayed, and sold. It promises to transform art discovery the way Shazam transformed music discovery.

ArtPI detects features and patterns using overarching visual style characteristics based in art theory and history. Built from the ground up specifically for art and drawing on centuries of artworks, ArtPI can find visually similar works, label styles and eras, recognize subject matter and artist, and find connections within a large collection.

For museums and other institutions, this means users can search their collection visually, exploring artworks in a more organic and intuitive way by looking for particular features. Curators can have the entire collection at their fingertips, finding unexpected ties and similarities that may inspire new ways to display or interpret works. For galleries, dealers, and auction houses, ArtPI can be used in conjunction with social media to uncover trends or predict market conditions.

“AI is often used to label images and objects in them, and ArtPI can also find works by subject matter,” explains Ahmed Elgammal, creator of ArtPI, Professor of Computer Science at Rutgers University, and founder of AI art startup Artrendex. “But you don’t always just want to find painting of dogs or putti. Sometimes you want a particular style of lighting or line. We’ve trained our algorithms to identify these elements and principles of art, based on foundational art historians’ approaches. By working with the visual elements instead of content labels or other metadata like artists or era, you can find art you didn’t know existed and see its connections with other works.”

ArtPI works even when images of an artwork are distorted, allowing for less-than-ideal source images or shots from a variety of angles. The AI model has been trained to correct for distortion, distilling a slanted or poorly lit image into information that can be compared to other works. This does away with the need for complicated photographic documentation, letting ArtPI work in a variety of less formal circumstances. Now, for example, you can see what work from a major exhibit won the hearts of Instagramers, even if they took a quick snapshot from the side.

AI can spot visual patterns, often better than humans. ArtPI employs neural networks to train its algorithms on images, using more than 250,000 artworks for its initial training phase. The more works encountered by the algorithm, the more nuanced the results. “Humans grasp context quickly, but AI is great at homing in on details and doing so for millions of examples in seconds or fraction of second,” says Elgammal. “Every search and every image added to ArtPI will contribute to its refinement, increasing its value over time.”

Philadelphia’s Barnes Foundation was one of the first institutions to adapt ArtPI. The Barnes first analyzed its extensive collection, which includes a great number of impressionist works. The first round of AI-powered analysis by different computer vision platforms generated some unexpected results: In one instance, the content of museum’s Renoirs were labeled as stuffed animals. However, once the museum’s works were inputted to ArtPI, some remarkable connections emerged.

“We thought this part of our process would be the hardest, yet it proved the easiest,” recalls Shelley Bernstein, the digital consultant on the Barnes project. “The visual similarities were so good, they were almost too good.” The problem: The Barnes collection had a wonderfully idiosyncratic way of displaying works together that originated with the collection’s founder. To recreate this approach online, the Barnes and ArtPI crafted a custom way to display results that better mirrors the real-life museum experience.

“Exploring art, be it for educational or valuation purposes, can be done in many ways, and AI is just another tool, another frame that helps us find new insights into the art we’re working with,” says Elgammal. “Being aware and optimized for the concepts of art, ArtPI is designed to be the infrastructure recognition engine for museums and the art market, in other words, the ‘Shazam for Art.’ I believe visual search and suggested connection can both democratize the discovery process for laypeople and renew the excitement curators, galleries, and auction houses feel as they consider the works in their domain.”