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 Update ai: US, Israel strike secret Natanz nuclear facility, no radiat

A joint strike by the US and Israel targeted Iran’s Natanz nuclear facility early Friday, Iranian authorities reported. The country’s atomic energy organisation said the Natanz enrichment complex, one of Iran’s most sensitive uranium sites, was hit but emphasized that no radioactive materials were released. 

The facility, officially named the Shahid Ahmadi Roshan plant, is located approximately 135 miles southeast of Tehran in central Iran. 

The statement, carried by Iran’s Tasnim news agency, condemned the attacks as criminal actions by the US and Israel, while reassuring that the site’s operations and surrounding areas remain secure and radiation levels are unaffected.

Safe conditions after Natanz attack

Iranian authorities reported that no radioactive materials were released following an attack on the Natanz nuclear facility, according to Tasnim. Officials confirmed that the population living near the site faces no danger.

The Natanz enrichment plant, located in central Iran, was previously targeted by Israel during the 12-day conflict between Iran and Israel in June 2025. Al Jazeera reported that the statement from Iran’s atomic energy organisation did not provide details on how Saturday’s strike occurred or what type of munitions were used.

Last week, an unidentified projectile struck the grounds near Iran’s only operating nuclear power plant – the Bushehr Nuclear Power Plant – but international and Iranian officials said the impact caused no damage or injuries. The International Atomic Energy Agency (IAEA) and Iranian authorities confirmed the strike, noting that the reactor and other critical infrastructure remained intact and that radiation levels were not affected.

The Bushehr Nuclear Power Plant is the country’s only operational nuclear power station, equipped with a Russian‑built pressurized‑water reactor that produces around 1,000 megawatts of electricity. That output is equivalent to enough power for many hundreds of thousands of homes, but it represents only a small fraction of Iran’s overall electricity supply, or roughly 1 % to 2 % of total national generation.

IAEA investigates reported attack on Natanz

The International Atomic Energy Agency (IAEA) said it is investigating reports of a strike on Iran’s Natanz nuclear facility. IAEA Director General Rafael Grossi emphasized the need for military restraint to prevent any potential nuclear accidents. The agency confirmed that Iran had notified it of the attack and that no rise in radiation levels was detected outside the site.

At the same time, Israeli Defense Minister Israel Katz signaled that military operations against Iran could escalate. Katz said the United States and Israel plan to increase the intensity of strikes targeting Iran’s nuclear and military infrastructure in the coming week. He added that operations carried out by the IDF and US forces against what he called the “Iranian terror regime” would rise significantly.

US officials have said one of the central aims of the military campaign it began with Israel on February  28 is to ensure Iran does not acquire a nuclear weapons capability. The White House has framed this objective as part of broader efforts to counter Tehran’s nuclear ambitions and reduce the risk of nuclear proliferation in the region, asserting that stopping a potential Iranian bomb remains a key justification for the joint operations.

🔗 Sumber: interestingengineering.com