MAROKO133 Eksklusif ai: ByteDance Introduces Astra: A Dual-Model Architecture for Autonomo

📌 MAROKO133 Hot ai: ByteDance Introduces Astra: A Dual-Model Architecture for Auto

The increasing integration of robots across various sectors, from industrial manufacturing to daily life, highlights a growing need for advanced navigation systems. However, contemporary robot navigation systems face significant challenges in diverse and complex indoor environments, exposing the limitations of traditional approaches. Addressing the fundamental questions of “Where am I?”, “Where am I going?”, and “How do I get there?”, ByteDance has developed Astra, an innovative dual-model architecture designed to overcome these traditional navigation bottlenecks and enable general-purpose mobile robots.

Traditional navigation systems typically consist of multiple, smaller, and often rule-based modules to handle the core challenges of target localization, self-localization, and path planning. Target localization involves understanding natural language or image cues to pinpoint a destination on a map. Self-localization requires a robot to determine its precise position within a map, especially challenging in repetitive environments like warehouses where traditional methods often rely on artificial landmarks (e.g., QR codes). Path planning further divides into global planning for rough route generation and local planning for real-time obstacle avoidance and reaching intermediate waypoints.

While foundation models have shown promise in integrating smaller models to tackle broader tasks, the optimal number of models and their effective integration for comprehensive navigation remained an open question.

ByteDance’s Astra, detailed in their paper “Astra: Toward General-Purpose Mobile Robots via Hierarchical Multimodal Learning” (website: https://astra-mobility.github.io/), addresses these limitations. Following the System 1/System 2 paradigm, Astra features two primary sub-models: Astra-Global and Astra-Local. Astra-Global handles low-frequency tasks like target and self-localization, while Astra-Local manages high-frequency tasks such as local path planning and odometry estimation. This architecture promises to revolutionize how robots navigate complex indoor spaces.

Astra-Global: The Intelligent Brain for Global Localization

Astra-Global serves as the intelligent core of the Astra architecture, responsible for critical low-frequency tasks: self-localization and target localization. It functions as a Multimodal Large Language Model (MLLM), adept at processing both visual and linguistic inputs to achieve precise global positioning within a map. Its strength lies in utilizing a hybrid topological-semantic graph as contextual input, allowing the model to accurately locate positions based on query images or text prompts.

The construction of this robust localization system begins with offline mapping. The research team developed an offline method to build a hybrid topological-semantic graph G=(V,E,L):

• V (Nodes): Keyframes, obtained by temporal downsampling of input video and SfM-estimated 6-Degrees-of-Freedom (DoF) camera poses, act as nodes encoding camera poses and landmark references.
• E (Edges): Undirected edges establish connectivity based on relative node poses, crucial for global path planning.
• L (Landmarks): Semantic landmark information is extracted by Astra-Global from visual data at each node, enriching the map’s semantic understanding. These landmarks store semantic attributes and are connected to multiple nodes via co-visibility relationships.

In practical localization, Astra-Global’s self-localization and target localization capabilities leverage a coarse-to-fine two-stage process for visual-language localization. The coarse stage analyzes input images and localization prompts, detects landmarks, establishes correspondence with a pre-built landmark map, and filters candidates based on visual consistency. The fine stage then uses the query image and coarse output to sample reference map nodes from the offline map, comparing their visual and positional information to directly output the predicted pose.

For language-based target localization, the model interprets natural language instructions, identifies relevant landmarks using their functional descriptions within the map, and then leverages landmark-to-node association mechanisms to locate relevant nodes, retrieving target images and 6-DoF poses.

To empower Astra-Global with robust localization abilities, the team employed a meticulous training methodology. Using Qwen2.5-VL as the backbone, they combined Supervised Fine-Tuning (SFT) with Group Relative Policy Optimization (GRPO). SFT involved diverse datasets for various tasks, including coarse and fine localization, co-visibility detection, and motion trend estimation. In the GRPO phase, a rule-based reward function (including format, landmark extraction, map matching, and extra landmark rewards) was used to train for visual-language localization. Experiments showed GRPO significantly improved Astra-Global’s zero-shot generalization, achieving 99.9% localization accuracy in unseen home environments, surpassing SFT-only methods.

Astra-Local: The Intelligent Assistant for Local Planning

Astra-Local acts as the intelligent assistant for Astra’s high-frequency tasks, a multi-task network capable of efficiently generating local paths and accurately estimating odometry from sensor data. Its architecture comprises three core components: a 4D spatio-temporal encoder, a planning head, and an odometry head.

The 4D spatio-temporal encoder replaces traditional mobile stack perception and prediction modules. It begins with a 3D spatial encoder that processes N omnidirectional images through a Vision Transformer (ViT) and Lift-Splat-Shoot to convert 2D image features into 3D voxel features. This 3D encoder is trained using self-supervised learning via 3D volumetric differentiable neural rendering. The 4D spatio-temporal encoder then builds upon the 3D encoder, taking past voxel features and future timestamps as input to predict future voxel features through ResNet and DiT modules, providing current and future environmental representations for planning and odometry.

The planning head, based on pre-trained 4D features, robot speed, and task information, generates executable trajectories using Transformer-based flow matching. To prevent collisions, the planning head incorporates a masked ESDF loss (Euclidean Signed Distance Field). This loss calculates the ESDF of a 3D occupancy map and applies a 2D ground truth trajectory mask, significantly reducing collision rates. Experiments demonstrate its superior performance in collision rate and overall score on out-of-distribution (OOD) datasets compared to other methods.

The odometry head predicts the robot’s relative pose using current and past 4D features and additional sensor data (e.g., IMU, wheel data). It trains a Transformer model to fuse information from different sensors. Each sensor modality is processed by a specific tokenizer, combined with modality embeddings and temporal positional embeddi…

Konten dipersingkat otomatis.

🔗 Sumber: syncedreview.com

📌 MAROKO133 Hot ai: We Could Hitch a Ride to Unknown Frontiers on Super-Fast Inter

Mysterious interstellar object 3I/ATLAS made its closest approach to Earth on December 19, coming within just 167 million miles.

Scientists have been closely monitoring the object — which is largely believed to be a natural comet and only the third of its kind to have been directly observed in the solar system — as it continued on its highly eccentric trajectory.

The encounter with Earth, however, turned out to be a bit of an anticlimax, as Harvard astronomer Avi Loeb, who has long championed the far-fetched theory that the object may be an alien spacecraft, lamented in a blog post titled “3I/ATLAS Ignores Earth.” Instead of doing something you might expect of aliens during their closest approach to Earth, it simply cruised on by.

While hopes that we were just visited by an alien race diminish even further, Loeb made an interesting pivot in a follow-up piece, proposing that other objects like 3I/ATLAS could be useful for our future attempts to explore beyond our solar system.

“The Voyager Golden Records, containing a time capsule of sounds, images, music, and messages from Earth, were attached to NASA’s Voyager 1 and 2 spacecraft, which are currently traveling out of the solar system,” he wrote. “These records serve as humanity’s message for any intelligent extraterrestrial life that might find them, essentially a ‘message in a bottle’ sent out to interstellar space.”

Thanks to their tremendous speeds, exotic visitors like the latest interstellar object could carry a human spacecraft out of the solar system on our behalf, Loeb argued.

By “riding 3I/ATLAS,” which is traveling at a speed of 37 miles a second, he argued that we could reach “interstellar space by the year ~10,000 CE instead of the year ~30,000 CE.”

Even though they launched almost half a century ago, Voyager 1 and Voyager 2 only recently reached the boundary of the heliosphere. To enter true interstellar space beyond the Oort Cloud, where objects are no longer bound to the gravitational effects of the Sun, it could take Voyager 1 another 28,000 years.

“The discovery of interstellar objects over the past decade offers new opportunities for humanity to send time capsules to interstellar space,” he concluded.

To attach the message to objects like 3I/ATLAS, Loeb proposed using a “high-power laser beam to engrave a message” or “design interceptor missions” to attach technological objects to their surface.

For its part, 3I/ATLAS is expected to get within just 33.3 million miles of Jupiter on March 16, 2026, offering us yet another opportunity to have a closer look using spacecraft positioned there.

Nobody knows when we’ll be able to observe the next interstellar object careening through our neck of the cosmic woods. But according to Loeb, we should seize the opportunity.

To “gain respect near the top of the food chain in the Milky Way galaxy,” Loeb argued that we must take matters into our own hands, and “endeavor to interstellar space.”

Loeb himself says he’d even jump to commit his own body to such a project.

He “would have loved to hitchhike 3I/ATLAS and let it carry my remains into interstellar space,” he wrote, “if offered the opportunity.”

More on 3I/ATLAS: Mysterious Interstellar Object Now Approaching Earth

The post We Could Hitch a Ride to Unknown Frontiers on Super-Fast Interstellar Objects Like 3I/ATLAS appeared first on Futurism.

🔗 Sumber: futurism.com