Interview with Vanessa Rosa

Note: The award for Best Art Direction is shared by two films: this film and “Dance of the Nain Rouge”

Synopsis: Little Martians are future beings who claim to be the creators of our reality. One of them wants to apologize to humanity.

IAG: Before we get into the film, tell us a little bit about your previous work and and interests, which I know is quite unique.

VR: I fell in love with painting when I was a kid, learning from my grandmother and mom, and spending countless hours creating imaginary worlds with my sister. At 10 years old, I started illustrating my mother's stories, which led my family and me into the publishing industry where I developed my skills as an illustrator and author.

My path took an unexpected turn in 2010 when I discovered street art in Rio de Janeiro. What started as painting murals that merged historical characters with local environments evolved into a deeper exploration of cultural exchange – from Portuguese tiles to Islamic geometry, Chinese porcelain to indigenous patterns. Street art became my passport to the world, leading to projects across four continents, including significant works like a large-scale mural for Pioneer Works in New York and a commission for UN Women.

I started to explore laser cutting in 2015, initially to make elaborate stencils for my murals. This technical curiosity opened doors to hackerspaces and creative technology collectives, where I discovered the fascinating intersection of art and technology. By 2020, I was fully immersed in exploring AI as a creative tool, developing 'Little Martians' – a sci-fi universe that merges my traditional art skills with AI techniques, featured in the Nvidia AI Art Gallery since March 2023. Throughout this evolution, my work has remained focused on storytelling, whether through children's books like the "Diana's World" series, immersive installations, or my latest experiments merging physical and digital media.

IAG: This film has won two of our awards—Best Art Direction and Best Original Song. Let’s start with the song. Where did the lyrics come from, and how did you create the music?

VR: I've been playing with lyrics for my Little Martians series for years, jotting down ideas and occasionally recording myself singing, despite never considering myself particularly musical. I always felt the project needed a musical dimension, but I wasn't sure how to bring it to life until Suno v3 came along. The lyrics themselves came together in a single afternoon, though I think they had been brewing in my subconscious for weeks. I worked with ChatGPT to refine the rhythm and rhyme structure, as those technical aspects of songwriting weren't my strong suit.

Testing the lyrics with Suno was a revelation. I crafted a very specific musical direction in my prompt: "bossa nova, violin, eastern European, classic, mellow, choro." After dozens of generations and careful tweaking, one version stood out magnificently. The process involved several rounds of editing and regenerating sections until everything clicked. What amazed me was how the finished song became the heartbeat of the animation – I found myself listening to it repeatedly, letting it guide my visualization of each scene.

What fascinates me most about this process is how it bridges the gap between imagination and execution. I've always heard the music of my stories in my head, but now I can actually bring those sounds into the world. It's not about becoming a musician – it's about having another color in my artistic palette, another way to make my characters and their world come alive.

IAG: Now let’s get into the visuals—which, I must say, are delightfully trippy. Technically speaking, how did you create them?

VR: It all started with my hands in clay, sculpting this character that had been living in my imagination. There's something special about beginning with a physical sculpture - it grounds the whole project in reality. Once I had my ceramic figure, I scanned it into the digital world, but that was just the beginning of its transformation. Using eden.art's LORA AI technique, I trained the AI to understand my character's essence, teaching it to create images with characters in any context.  

The animation phase brought the character to life, employing NVIDIA’s Audio2face to animate the facial expressions of the 3D model, infusing it with emotional nuance. This was followed by the creation of animated 3D renders, which were artistically transformed using eden.art’s vid2vid animatediff AI model. The integration of AI transformed copyright-free footage from Pexels, along with the innovative use of real2real and img2vid techniques on top of the LORA model, introduced dynamic backgrounds and scenes.

The final assembly of the animation was a meticulous process, utilizing Runway’s Gen2 and greenscreen model for more visual effects and Ebsynth for precise motion control. The editing process was conducted in After Effects and Premiere Pro, where the various layers of visuals and music were seamlessly blended. 

What excites me most is how this blend of physical craft and digital innovation opens up entirely new possibilities for storytelling - it's like having an entire art studio and a team of digital assistants at your fingertips.