AI Videos
  • All Technology
  • AI
  • Autonomy
  • B2B Growth
  • Big Data
  • BioTech
  • ClimateTech
  • Consumer Tech
  • Crypto
  • Cybersecurity
  • DevOps
  • Digital Marketing
  • Ecommerce
  • EdTech
  • Enterprise
  • FinTech
  • GovTech
  • Hardware
  • HealthTech
  • HRTech
  • LegalTech
  • Nanotech
  • PropTech
  • Quantum
  • Robotics
  • SaaS
  • SpaceTech
AllNewsDealsSocialBlogsVideosPodcastsDigests

AI Pulse

EMAIL DIGESTS

Daily

Every morning

Weekly

Sunday recap

NewsDealsSocialBlogsVideosPodcasts
AIVideosYou’ll Never Look At Chocolate TV Ads The Same Way Again
AI

You’ll Never Look At Chocolate TV Ads The Same Way Again

•November 14, 2025
0
Two Minute Papers
Two Minute Papers•Nov 14, 2025

Why It Matters

The adaptive octree technique dramatically lowers the cost of high‑fidelity liquid effects, enabling advertisers and VFX studios to create realistic splashes that were previously impractical, and it points to a near‑future where such simulations could run in real time.

Summary

The video explains a breakthrough in computer‑generated fluid dynamics that could finally make the impossible‑looking chocolate‑and‑caramel splashes in TV ads look authentic. It centers on a five‑year‑old research paper by Ryoichi Ando, advised by Chris Batty, which introduces a novel “staggered octree Poisson discretization” for free‑surface liquids. By adapting the simulation grid—refining it only where the splash is most violent and coarsening it elsewhere—the method sidesteps the massive computational load of a uniform billion‑point 3‑D grid.

Key technical insights include the use of octrees to create an adaptive hierarchy of boxes, and a second‑order pressure solver that remains smooth across octree T‑junctions without resorting to costly Voronoi diagram constructions. The approach delivers high‑resolution detail where needed while keeping overall calculations tractable, though each frame still requires 1.5‑3 minutes of processing. The presenter demonstrates wireframe visualizations, a fish‑swim scene, and side‑by‑side comparisons that highlight the elimination of artifacts that plagued earlier methods.

Notable moments feature the creator’s praise—“Genius!”—and the explicit acknowledgment that the technique has been largely unnoticed despite its potential. The video showcases the algorithm’s ability to render extreme adaptivity, likening its robustness to a “Toyota that just refuses to break.” It also stresses that the method is not yet real‑time, but the runtime is acceptable for high‑budget advertising or cinematic VFX where frame‑by‑frame perfection outweighs speed.

The implications are significant for the advertising and visual‑effects industries. By making photorealistic liquid simulations feasible without prohibitive compute costs, brands can produce more convincing product shots—like caramel perfectly coating a chocolate bar—without expensive physical rigs. Moreover, the research underscores the importance of surfacing overlooked academic work, suggesting that wider adoption could accelerate the path toward real‑time, physically accurate fluid rendering.

Original Description

❤️ Check out Weights & Biases and sign up for a free demo here: https://wandb.me/papers
📝 The paper "A practical octree liquid simulator with adaptive surface resolution" is available here:
https://cs.uwaterloo.ca/~c2batty/papers/Ando2020/Ando2020.pdf
Sources:
https://www.youtube.com/watch?v=kdt5Cs1VYJA
https://www.youtube.com/watch?v=YmmSDZ6dBdY
https://www.youtube.com/shorts/FVIDRU9-FW8
https://www.youtube.com/watch?v=gNZtx3ijjpo&pp=ygUHb2N0cmVlcw%3D%3D
https://www.youtube.com/shorts/1Euba1QvhW0
https://www.youtube.com/shorts/k2P9yWSMaXE
https://www.youtube.com/watch?v=Z5qbxQI6dgw
https://www.youtube.com/watch?v=laoGmqNtUMI
📝 My paper on simulations that look almost like reality is available for free here:
https://rdcu.be/cWPfD
Or this is the orig. Nature Physics link with clickable citations:
https://www.nature.com/articles/s41567-022-01788-5
🙏 We would like to thank our generous Patreon supporters who make Two Minute Papers possible:
Benji Rabhan, B Shang, Christian Ahlin, Gordon Child, Juan Benet, Michael Tedder, Owen Skarpness, Richard Sundvall, Steef, Taras Bobrovytsky, Tybie Fitzhugh, Ueli Gallizzi
If you wish to appear here or pick up other perks, click here: https://www.patreon.com/TwoMinutePapers
My research: https://cg.tuwien.ac.at/~zsolnai/
X/Twitter: https://twitter.com/twominutepapers
Thumbnail design: Felícia Zsolnai-Fehér - http://felicia.hu
0

Comments

Want to join the conversation?

Loading comments...