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

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

📌 TOPINDIATOURS Hot ai: Trio wins 2025 Nobel Prize in Chemistry for groundbreaking

The 2025 Nobel Prize in Chemistry has been awarded to Susumu Kitagawa, Richard Robson, and Omar M. Yaghi for their groundbreaking work on metal-organic frameworks (MOFs) on Wednesday.

MOFs are a revolutionary class of materials whose molecular structures contain “rooms for chemistry.”

The Royal Swedish Academy of Sciences said the award recognizes the trio “for the development of metal-organic frameworks,” which are crystalline materials made by linking metal ions with organic molecules to form highly porous structures.

These frameworks can trap, store, and manipulate gases and molecules, offering vast potential in tackling global sustainability challenges. By designing structures with enormous internal surface areas, MOFs allow gases such as carbon dioxide, methane, and water vapor to flow in and out through tiny cavities.

This property enables them to perform remarkable functions, from capturing greenhouse gases and purifying water to catalyzing chemical reactions and storing hydrogen fuel.

Scientists describe these materials as “molecular architecture with purpose-built rooms,” capable of hosting new and tailored chemistry within their structures.

Origins and development of MOFs

The innovation traces back to 1989 when Richard Robson experimented with assembling copper ions and complex organic molecules into spacious crystalline frameworks.

Early structures were unstable, but his work inspired further research. In the 1990s, Susumu Kitagawa demonstrated that these frameworks could absorb and release gases, showing their flexibility and potential.

Omar Yaghi later engineered the first exceptionally stable MOFs and introduced rational design principles. These principles allow chemists to fine-tune MOFs for specific properties, making them suitable for a wide range of applications.

According to Heiner Linke, Chair of the Nobel Committee for Chemistry, “Metal-organic frameworks have enormous potential, bringing previously unforeseen opportunities for custom-made materials with new functions.”

Since these early discoveries, chemists have synthesized tens of thousands of MOFs, applying them to carbon capture, filtering pollutants, harvesting water from desert air, and converting chemicals efficiently.

Impact on Sustainability and Materials Science

The 2025 Chemistry laureates’ work has transformed materials science. MOFs provide tools for addressing some of the world’s most pressing environmental and energy challenges.

Their ability to host and manipulate molecules at the nanoscale opens possibilities for sustainable solutions in energy storage, pollution reduction, and water purification.

The development of MOFs has also paved the way for research into advanced chemical processes. Scientists can now design frameworks for specific tasks, creating materials with properties that were once impossible.

This customizability has the potential to change the way industries approach chemical engineering and environmental solutions.

Physics Nobel Prize highlights quantum breakthroughs

On Tuesday, the Royal Swedish Academy of Sciences awarded the 2025 Nobel Prize in Physics to John Clarke, Michel H. Devoret, and John M. Martinis for work on quantum mechanical tunneling.

The award recognizes their research “for the discovery of macroscopic quantum mechanical tunnelling and energy quantisation in an electric circuit.”

Their experiments demonstrated that quantum mechanical properties can exist on a macroscopic scale.

The academy said their work has “provided opportunities for developing the next generation of quantum technology, including quantum cryptography, quantum computers, and quantum sensors.”

Quantum computing uses principles of quantum mechanics to perform calculations faster than conventional computers. The chair of the Nobel Committee for Physics noted that quantum mechanics is the “foundation of all digital technology.”

🔗 Sumber: interestingengineering.com