New approaches, new services, new technologies

El Mercado (Podcast)

[Since 07.2024]

Objective:

To talk to different stakeholders who are part of the marketing, audiovisual, and creative industries. It’s an open and honest conversation, with no payments or compensations, like the kind of chats neighbors have when they bump into each other, like the ones we have when we exchange ideas… at El Mercado.

Context:

In the marketing, audiovisual, and creative industries, spontaneity and authenticity are often overshadowed by fear, secrecy, and egos. This environment hinders open dialogue. There is a need to foster honest conversations that reveal fresh ideas and tackle the evolving challenges and opportunities our industries face.

Cleaner and Greener Chroma Experiment

[08.2024]

Objective:

Exploring AI’s potential in conjunction with chroma screen while maintaining a high-quality short-form content!.

Context:

The project delivers a captivating audiovisual piece that merges AI-generated visuals with chroma key technology. Utilizing advanced AI models, it crafts both music and imagery, while product close-ups are dynamically integrated through chroma techniques. This approach ensures a visually striking and innovative presentation. Select product manipulations and the final assembly are refined using professional video editing software.

Specs:

  • Length: 15 Sec.
  • Type: Snackable content
  • Main Techs: AI-generated elements (music and images)
  • Chroma Screen: Green Screen
  • Objects Manipulation: Adobe After Effects
  • Assembly: Adobe Premiere
Play Video

Results:
This innovative experiment redefines AI’s capabilities in conjunction with Chroma Screen techniques. Crafted to elevate a cleaning products brand, the content is immersive, visually striking, and pioneering.

While leveraging advanced AI to produce high-quality visuals, the experiment is grounded in a deep understanding of traditional audiovisual techniques. This expertise bridges the gap between current technological limitations and the desired outcome, ensuring a polished final product.

Realistic Visual Effects: Demonstrates AI’s enhanced ability to create lifelike sets and visuals, boosting fluidity and reinforcing brand messaging.

Chroma Screen Integration: Fuses cutting-edge AI technology with chroma screen techniques, delivering a cohesive and engaging viewer experience.

Elevated Impact: The 15-second format is optimized for maximum impact, enhancing social media engagement and strengthening the brand’s visual identity.

Cool and Liquid Experiment

[08.2024]

Objective:

To push AI’s potential in liquid and beverages visuals, as well as deep-face generation, while maintaining high-quality short-form content.

Context:

The project generates a snackable audiovisual piece that explores liquid visuals and deep-face technology generated by AI. It utilizes state of the art AI models to create every element -music, audio, images and voice-overs-, ensuring a captivating and cutting-edge presentation. The final assembly is crafted in video editing software.

Specs:

  • Length: 15 Sec.
  • Type: Snackable content
  • Main Techs: AI-generated elements (music, audio, voice-over, images, deep-face, liquid videos)
  • DeepFace: Cutting-Edge (non-mainstream AI models)
  • Assembly: Adobe Premiere
Play Video

Results:
This experiment pushes AI’s boundaries in generating liquid and beverage visuals while seamlessly integrating deep-face technology. We craft an immersive, visually stunning and innovative content designed to elevate a soda brand.

Producing high-quality material based on AI, still relies on a profound knowledge of traditional audiovisual resources, which helps overcome the technological shortcomings present in current technology.

Realistic Liquid Effects: Successfully demonstrates AI’s ability to generate realistic liquid visuals, enhancing fluidity and brand messaging.

Seamless Deep-Face Integration: Effectively incorporates cutting-edge deep-face technology, blending smoothly with liquid visuals for a cohesive experience.

Engaging Audio Sync: Harmoniously combines AI-generated music, audio, and voice-over with visuals.

Elevated Impact: The 15-second format delivers a sharp, impactful showcase of AI usage, particularly boosting social media interactions while elevating the brand’s visual identity.

A Coffee Brand Experiment

[07.2024]

Objective:

Create a high quality snackable audiovisual piece, exclusively using AI-generated elements.

Context:

The project leverages cutting-edge AI technology to produce all aspects of the content, ensuring a seamless and innovative presentation. The final assembly is executed in a video editing software.

Process:

  • Music (Foley Sounds): Generated by AI
  • Audio: Produced by AI
  • Voice-over: Created with AI
  • Images: Designed using AI
  • Videos: Generated through AI
  • Assembly: Adobe Premiere
Play Video

Results:
This experiment demonstrates the impressive capabilities of AI in crafting short-form content with fluidity, visual appeal, and narrative coherence. By leveraging AI for music, audio, voice-over, imagery, and video, the project achieves a harmonious and engaging presentation.

However, one current limitation is the challenge of incorporating longer micro-scenes while maintaining video quality and narrative consistency. This constraint is one reason the clip is set to a precise 15 seconds, allowing for a focused and impactful delivery while highlighting the potential of AI in content creation.

This approach not only showcases the advancements in AI-driven production but also sets a benchmark for future innovations in the field.

The GABO Experiment

[06.2024]

Objective:

[1] To evaluate the state of the art in deepface technology.

[2] To evaluate the state of the art in deepvoice technology.

[3] To match these technologies and evaluate their professional and commercial use.

[4] To reduce to a minimum the post-production of a video clip.

Context:

[1] At the time of the experiment there is no known mainstream-commercial deepface service.

[2] The development team of The Vecinos used the most advanced open-source model for deepface.

[3] The team used a commercial deepvoice service.
Play Video
Results:
[1] At the current time the deepface quality depends on the “base clip” that is passed to the deep learning model.
[2] In many situations, especially when dealing with old or low quality material, this can constitute an issue hard to overcome.
[3] Still, the advancements and pace of improvement are extremely promising in the near term.
[4] With some light post-production, a vintage-looking or other artistic alternatives can compensate the current technological handicap. This solution was implemented in the experiment.
[5] The deepvoice technology is commercially ready at a professional grade.