BREAKING NEWS
LATEST POSTS
-
NVidia Physical AI – Collection of commercial-grade datasets for physical AI developers
https://huggingface.co/collections/nvidia/physical-ai-67c643edbb024053dcbcd6d8
🔹 15TB of high-quality, standardized synthetic data
🔹 320,000+ trajectories for robotics training
🔹 1,000+ OpenUSD assets, including a SimReady collection
-
RGBAvatar – Reduced Gaussian Blendshapes for Online Modeling of Head Avatars
https://gapszju.github.io/RGBAvatar
A method for reconstructing photorealistic, animatable head avatars at speeds sufficient for on-the-fly reconstruction. Unlike prior approaches that utilize linear bases from 3D morphable models (3DMM) to model Gaussian blendshapes, our method maps tracked 3DMM parameters into reduced blendshape weights with an MLP, leading to a compact set of blendshape bases.
https://github.com/gapszju/RGBAvatar
-
Robert Legato joins Stability AI as Chief Pipeline Architect
https://stability.ai/news/introducing-our-new-chief-pipeline-architect-rob-legato
“Joining Stability AI is an incredible opportunity, and I couldn’t be more excited to help shape the next era of filmmaking,” said Legato. “With dynamic leaders like Prem Akkaraju and James Cameron driving the vision, the potential here is limitless. What excites me most is Stability AI’s commitment to filmmakers—building a tool that is as intuitive as it is powerful, designed to elevate creativity rather than replace it. It’s an artist-first approach to AI, and I’m thrilled to be part of it.”
-
WSL – Windows Subsystem for Linux
https://learn.microsoft.com/en-us/windows/wsl/install
The Windows Subsystem for Linux (WSL) is a feature of the Windows operating system that enables you to run a Linux file system, along with Linux command-line tools and GUI apps, directly on Windows, alongside your traditional Windows desktop and apps.
https://ubuntu.com/desktop/wsl
-
Apple TV+ Reportedly Loses $1 Billion on Streaming a Year
https://www.indiewire.com/news/business/apple-tv-reportedly-loses-1-billion-streaming-1235110323
But it supposedly also has 45 million subscribers, and with $124.3 million in revenue.
Per the company’s most recent earnings, the three months ending in January saw Apple bring in $124.3 billion in revenue, $26.3 billion of which came from Services, a record for the division. That’s just for one quarter. For the year, Services brought in more than $96 billion. It can afford to absorb a billion dollars in losses.https://spyglass.org/apple-tv-plus-strategy
It Wasn’t the Apple TV+ Spend, It Was the Apple TV+ Strategy
-
VillageRoadShow production studio files for bankruptcy
Village Roadshow (prod company/financier: Wonka, the Matrix series, and Ocean’s 11) has filed for bankruptcy.
It’s a rough indicator of where we are in 2025 when one of the last independent production companies working with the studios goes under.
Here’s their balance sheet:
$400 M in library value of 100+ films (89 of which they co-own with Warner Bros.)
$500 M – $1bn total debt
$1.4 M in debt to WGA, whose members were told to stop working with Roadshow in December
$794 K owed to Bryan Cranston’s prod company
$250 K owed to Sony Pictures TV
$300 K/month overhead
The crowning expense that brought down this 36-year-old production company is the $18 M in (unpaid) legal fees from a lengthy and currently unresolved arbitration with their long-time partner Warner Bros, who they’ve had a co-financing arrangement since the late 90s.
Roadshow sued when WBD released their Matrix Resurrections (2021) film in theaters and on Max simultaneously, causing Roadshow to withhold their portion of the $190 M production costs.
Due to mounting financial pressures, Village Roadshow’s CEO, Steve Mosko, a veteran film and TV exec, left the company in January.
Now, this all falls on the shoulders of Jim Moore, CEO of Vine, an equity firm that owns Village Roadshow, as well as Luc Besson’s prod company EuropaCorp.
FEATURED POSTS
-
MiniMax-Remover – Taming Bad Noise Helps Video Object Removal Rotoscoping
https://github.com/zibojia/MiniMax-Remover
MiniMax-Remover is a fast and effective video object remover based on minimax optimization. It operates in two stages: the first stage trains a remover using a simplified DiT architecture, while the second stage distills a robust remover with CFG removal and fewer inference steps.
