MAROKO133 Breaking ai: ByteDance Introduces Astra: A Dual-Model Architecture for Autonomou

📌 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: Entire B-52 bomber fleet can be nuclear capable if ordered:

The U.S. can restore nuclear capability to the entire B-52 bomber fleet, according to its Air Force Global Strike Command. The statement came as the New START treaty, which imposed arms control restrictions on the US and Russia, has recently expired.

The Air Force Global Strike Command stated that if ordered, it can restore nuclear capability to the B-52 fleet and can load more warheads onto Minuteman III intercontinental ballistic missiles (ICBM), according to a report.

The treaty also limited all Russian deployed intercontinental-range nuclear weapons.

Dual-capable long-range strike platforms

The New START treaty also limited all Russian deployed intercontinental-range nuclear weapons, including every Russian nuclear warhead that is loaded onto an intercontinental-range ballistic missile that can reach the United States in approximately 30 minutes. It also limits the deployed Avangard and the under-development Sarmat, the two most operationally available of the Russian Federation’s new long-range nuclear weapons that can reach the United States.

“The conclusion of New START allows us to streamline our focus and dedicate more resources to our core mission: ensuring a safe, secure, and effective nuclear deterrent,” an AFGSC spokesperson told The War Zone. “This managed transition enhances our operational readiness and our ability to respond to the nation’s call.”

“Although we will not comment on the posturing of our forces, Air Force Global Strike Command both maintains the capability and training to MIRV the Minuteman III ICBM force and convert its entire B-52 fleet into dual-capable long-range strike platforms if directed by the President,” added the spokesperson.

Treaty placed limits on all of Russia’s intercontinental-range nuclear weapons

The Treaty between the United States of America and the Russian Federation on Measures for the Further Reduction and Limitation of Strategic Offensive Arms, also known as the New START Treaty, enhances U.S. national security by placing verifiable limits on all Russian deployed intercontinental-range nuclear weapons.

The treaty offered each party the flexibility to determine for itself the structure of its forces, subject to the central limits. The New START Treaty gave the United States flexibility to deploy and maintain U.S. strategic nuclear forces in a way that best serves U.S. national security interests.

The U.S. Air Force Global Strike Command is ready to restore nuclear capability to the entire fleet of B-52 bombers, which can perform strategic attack, close-air support, air interdiction, offensive counter-air, and maritime operations.

During Desert Storm, B-52s delivered 40 percent of all the weapons dropped by coalition forces. It is highly effective when used for ocean surveillance and can assist the U.S. Navy in anti-ship and mine-laying operations. In two hours, two B-52s can monitor 140,000 square miles (364,000 square kilometers) of ocean surface.

It’s also claimed that all B-52s can be equipped with two electro-optical viewing sensors, a forward-looking infrared, and advanced targeting pods to augment targeting, battle assessment, and flight safety, further improving their combat ability. Pilots wear night vision goggles, or NVGs, to enhance their vision during night operations. Night vision goggles provide greater safety during night operations by increasing the pilot’s ability to visually clear terrain, increasing the peacetime and combat situational awareness of the aircrew, and improving their ability to visually acquire other aircraft.

🔗 Sumber: interestingengineering.com