📌 TOPINDIATOURS Breaking ai: Adobe Research Unlocking Long-Term Memory in Video Wo

Video world models, which predict future frames conditioned on actions, hold immense promise for artificial intelligence, enabling agents to plan and reason in dynamic environments. Recent advancements, particularly with video diffusion models, have shown impressive capabilities in generating realistic future sequences. However, a significant bottleneck remains: maintaining long-term memory. Current models struggle to remember events and states from far in the past due to the high computational cost associated with processing extended sequences using traditional attention layers. This limits their ability to perform complex tasks requiring sustained understanding of a scene.

A new paper, “Long-Context State-Space Video World Models” by researchers from Stanford University, Princeton University, and Adobe Research, proposes an innovative solution to this challenge. They introduce a novel architecture that leverages State-Space Models (SSMs) to extend temporal memory without sacrificing computational efficiency.

The core problem lies in the quadratic computational complexity of attention mechanisms with respect to sequence length. As the video context grows, the resources required for attention layers explode, making long-term memory impractical for real-world applications. This means that after a certain number of frames, the model effectively “forgets” earlier events, hindering its performance on tasks that demand long-range coherence or reasoning over extended periods.

The authors’ key insight is to leverage the inherent strengths of State-Space Models (SSMs) for causal sequence modeling. Unlike previous attempts that retrofitted SSMs for non-causal vision tasks, this work fully exploits their advantages in processing sequences efficiently.

The proposed Long-Context State-Space Video World Model (LSSVWM) incorporates several crucial design choices:

1. Block-wise SSM Scanning Scheme: This is central to their design. Instead of processing the entire video sequence with a single SSM scan, they employ a block-wise scheme. This strategically trades off some spatial consistency (within a block) for significantly extended temporal memory. By breaking down the long sequence into manageable blocks, they can maintain a compressed “state” that carries information across blocks, effectively extending the model’s memory horizon.
2. Dense Local Attention: To compensate for the potential loss of spatial coherence introduced by the block-wise SSM scanning, the model incorporates dense local attention. This ensures that consecutive frames within and across blocks maintain strong relationships, preserving the fine-grained details and consistency necessary for realistic video generation. This dual approach of global (SSM) and local (attention) processing allows them to achieve both long-term memory and local fidelity.

The paper also introduces two key training strategies to further improve long-context performance:

• Diffusion Forcing: This technique encourages the model to generate frames conditioned on a prefix of the input, effectively forcing it to learn to maintain consistency over longer durations. By sometimes not sampling a prefix and keeping all tokens noised, the training becomes equivalent to diffusion forcing, which is highlighted as a special case of long-context training where the prefix length is zero. This pushes the model to generate coherent sequences even from minimal initial context.
• Frame Local Attention: For faster training and sampling, the authors implemented a “frame local attention” mechanism. This utilizes FlexAttention to achieve significant speedups compared to a fully causal mask. By grouping frames into chunks (e.g., chunks of 5 with a frame window size of 10), frames within a chunk maintain bidirectionality while also attending to frames in the previous chunk. This allows for an effective receptive field while optimizing computational load.

The researchers evaluated their LSSVWM on challenging datasets, including Memory Maze and Minecraft, which are specifically designed to test long-term memory capabilities through spatial retrieval and reasoning tasks.

The experiments demonstrate that their approach substantially surpasses baselines in preserving long-range memory. Qualitative results, as shown in supplementary figures (e.g., S1, S2, S3), illustrate that LSSVWM can generate more coherent and accurate sequences over extended periods compared to models relying solely on causal attention or even Mamba2 without frame local attention. For instance, on reasoning tasks for the maze dataset, their model maintains better consistency and accuracy over long horizons. Similarly, for retrieval tasks, LSSVWM shows improved ability to recall and utilize information from distant past frames. Crucially, these improvements are achieved while maintaining practical inference speeds, making the models suitable for interactive applications.

The Paper Long-Context State-Space Video World Models is on arXiv

The post Adobe Research Unlocking Long-Term Memory in Video World Models with State-Space Models first appeared on Synced.

🔗 Sumber: syncedreview.com

📌 TOPINDIATOURS Eksklusif ai: Hydrogen-electric aircraft inches closer toward indu

The aviation sector across the world is undergoing a major shift toward cleaner technologies, and Beyond Aero has taken an important step forward in this transition.

The Paris-headquartered company recently completed the Preliminary Design Review (PDR) of its hydrogen-electric business aircraft.

The step marks a critical phase where the project moves from early concepts into more detailed engineering and development.

Aircraft’s overall design is mature enough to proceed with toward production

This milestone indicates that the aircraft’s overall design is mature enough to proceed with confidence toward production and certification, according to the company.

At the core of Beyond Aero’s innovation is its use of hydrogen-electric propulsion, a system that generates power through hydrogen fuel cells instead of traditional jet fuel combustion. This approach produces only water as a by-product, making it a promising solution for reducing aviation emissions. By focusing on business aviation, a segment known for its relatively high carbon footprint per passenger, the company aims to make a meaningful impact on sustainability while demonstrating the viability of hydrogen-powered flight.

“The Preliminary Design Review confirms that the aircraft configuration and its major systems — propulsion, hydrogen storage, aerodynamics and avionics — have reached the level of maturity required to support a certifiable architecture,” said Luiz Oliveira, chief engineer at Beyond Aero.

“With this milestone completed, the program moves on schedule into detailed design and verification of the aircraft’s integrated systems.”

Hydrogen-electric propulsion

A notable aspect of the project is its strong emphasis on safety and regulatory compliance. Beyond Aero is designing its aircraft in line with established certification standards typically applied to commercial aircraft, ensuring that the final product meets rigorous aviation safety requirements. Early collaboration with regulators such as the European Union Aviation Safety Agency helps streamline the certification pathway and reduces potential delays later in development.

“The completion of the Preliminary Design Review demonstrates that a certifiable hydrogen-powered business aircraft is achievable,” said Eloa Guillotin, Chief Executive Officer of Beyond Aero.

“Our objective is to develop a new business aircraft tailored to the constraints of hydrogen-electric propulsion, while meeting the performance, safety, and operational standards expected in business aviation.”

The company’s progress is supported by extensive testing and validation efforts. These include successful trials of smaller-scale prototypes as well as the development of advanced ground testing systems to evaluate the performance of the propulsion technology. Such a methodical approach allows engineers to refine the system based on real-world data, improving reliability and reducing technical risks before full-scale production.

Beyond Aero is also working to build the necessary ecosystem for hydrogen aviation. This includes partnerships with industry players, airport operators, and energy providers to ensure that hydrogen infrastructure will be available when the aircraft enters service. By addressing both the aircraft design and the supporting environment, the company is taking a comprehensive approach to bringing hydrogen-powered aviation closer to reality.

🔗 Sumber: interestingengineering.com