MAROKO133 Update ai: How a 2,000-year-old Roman tombstone ended up in a New Orleans backya

📌 MAROKO133 Eksklusif ai: How a 2,000-year-old Roman tombstone ended up in a New O

A 2,000-year-old Roman headstone that once belonged to a sailor in the praetorian fleet of ancient Rome has been discovered in a backyard in New Orleans, tracing an extraordinary journey from a war-torn Italian museum to a quiet Louisiana neighborhood.

The marble slab, inscribed in Latin and dating back to the second century CE, was identified earlier this year after homeowners stumbled upon it while clearing their yard.

A discovery that started with curiosity

The discovery was made by Daniella Santoro, an anthropologist at Tulane University, and her husband, Aaron Lorenz, who owns a historic home at 1106 Cambronne Street in New Orleans’ Carrollton neighborhood. According to the Preservation Resource Center of New Orleans (PRCNO), the couple came across an unusual flat marble slab with carved inscriptions that appeared to be in Latin while clearing undergrowth in their backyard.

Concerned that it might indicate a forgotten cemetery beneath their home, Santoro contacted Dr. D. Ryan Gray, an anthropologist at the University of New Orleans. After examining photos of the stone, Gray shared them with international colleagues, including Univ.-Prof. Harald Stadler from the University of Innsbruck and Dr. Susann S. Lusnia, Associate Professor of Classical Studies at Tulane University.

Their findings revealed that the inscription was a Roman funerary marker for a sailor named Sextus Congenius Verus, a soldier in the praetorian fleet Misenensis who lived 42 years and served 22. Further research confirmed that a stone matching this description had been missing from the city museum in Civitavecchia, Italy, since World War II.

From a war-torn museum to a Louisiana garden

The museum in Civitavecchia, located northwest of Rome, was heavily bombed by Allied forces between 1943 and 1944, resulting in the destruction of much of its collection. The museum did not reopen until 1970. Researchers concluded that the tombstone was likely lost during this chaotic period.

Once the artifact’s origins were confirmed, Santoro, Lusnia, and Gray worked with the Antiquities Coalition and the FBI’s Art Crime Team to facilitate its repatriation. “They helpfully agreed to pick up the stone and keep it in custody while the repatriation process began,” Gray said in his Preservation in Print write-up.

Still, one major mystery remained. How did the artifact make its way from Italy to a backyard in New Orleans?

A WWII soldier’s relic

The answer came days later when Erin Scott O’Brien, the former owner of the Cambronne Street house, saw a news report about the discovery. “We were in shock. We could not believe it,” O’Brien said. She recalled placing the tablet in her backyard 21 years ago, believing it to be just an art piece she inherited from her grandparents.

O’Brien’s maternal grandfather, Charles Paddock Jr., served in Italy during World War II and married his wife, Adele, there in 1946 before returning to New Orleans. The couple kept the marble tablet in a display case in their Gentilly home until their deaths in the 1980s. Family members had no idea of its historical significance or origin.

While it remains unclear whether Paddock bought the stone, was gifted it, or brought it home as a wartime souvenir, experts now believe it likely entered his possession during or shortly after the war, when many Italian cultural artifacts were displaced or sold amid the chaos.

Returning to its rightful home

With its history finally traced, the headstone of Sextus Congenius Verus is now under the custody of the FBI’s Art Crime Team, awaiting repatriation to Italy. “It’s amazing,” O’Brien said. “It’s wonderful that it’s going back to where it belongs.”

Dr. Lusnia, who personally visited the museum in Civitavecchia to confirm the artifact’s identity, said the museum staff were “excited to welcome it back” and hope to celebrate its return.

For Gray, the case was an extraordinary example of how local curiosity can lead to global revelations. “While we may never know exactly how Sextus Congenius Verus’ tombstone ended up in New Orleans,” he said, “we do know that the item is now safe and on its way to being properly displayed.”

From a Roman port city devastated by war to a New Orleans garden, the ancient headstone’s journey underscores how history can lie quietly beneath our feet, waiting, sometimes for millennia, to be rediscovered.

🔗 Sumber: interestingengineering.com

📌 MAROKO133 Update ai: Adobe Research Unlocking Long-Term Memory in Video World Mo

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