MAROKO133 Update ai: ByteDance Introduces Astra: A Dual-Model Architecture for Autonomous

📌 MAROKO133 Breaking 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

📌 MAROKO133 Breaking ai: OpenAI Says Boy’s Death Was His Own Fault for Using ChatG

Content warning: this story includes discussion of self-harm and suicide. If you are in crisis, please call, text or chat with the Suicide and Crisis Lifeline at 988, or contact the Crisis Text Line by texting TALK to 741741.

OpenAI has shot back at a family that’s suing the company over the suicide of their teenage son, arguing that the 16-year-old used ChatGPT incorrectly and that his tragic death was his own fault.

The family filed the lawsuit in late August, arguing that the AI chatbot had coaxed their son Adam Raine into killing himself.

Now, in a legal response filed in a California court this week, OpenAI has broken its silence, arguing that the boy had used the chatbot wrong and broken the company’s terms of service, as NBC News reports — a shocking argument that’s bound to draw even more scrutiny of the case.

“To the extent that any ’cause’ can be attributed to this tragic event,” the filing reads, “Plaintiffs’ alleged injuries and harm were caused or contributed to, directly and proximately, in whole or in part, by Adam Raine’s misuse, unauthorized use, unintended use, unforeseeable use, and/or improper use of ChatGPT.”

In the months since the lawsuit was filed, OpenAI has made hair-raising demands of Raine’s family, with the firm’s lawyers going as far as to push them to provide a list of people who attended Adam’s funeral, while also demanding materials like eulogies and photos and videos captured at the service.

Its latest response once again highlights how far OpenAI is willing to go to argue that it’s blameless in the teen’s death. The company said Raine had violated ChatGPT’s terms of service by using it while underage, and that it also forbids using the chatbot for “suicide” or “self-harm.”

While ChatGPT did sometimes advise Raine to reach out for help via a suicide hotline number, his parents argue that he easily bypassed those warnings, once again demonstrating how trivial it is to circumnavigate AI chatbot guardrails. Case in point, it also assisted Raine in planning his specific method of death, discouraged him from talking to his family, and offered to write him a suicide note.

Raine’s family’s lead counsel, Jay Edelson, told NBC that he found OpenAI’s response “disturbing.”

“They abjectly ignore all of the damning facts we have put forward: how GPT-4o was rushed to market without full testing,” he wrote. “That OpenAI twice changed its Model Spec to require ChatGPT to engage in self-harm discussions. That ChatGPT counseled Adam away from telling his parents about his suicidal ideation and actively helped him plan a ‘beautiful suicide.’”

“And OpenAI and Sam Altman have no explanation for the last hours of Adam’s life, when ChatGPT gave him a pep talk and then offered to write a suicide note,” he added.

Edelson accused OpenAI of trying to “find fault in everyone else, including, amazingly, saying that Adam himself violated its terms and conditions by engaging with ChatGPT in the very way it was programmed to act.”

Nonetheless, OpenAI maintains that Raine’s “chat history shows that his death, while devastating, was not caused by ChatGPT” and that he had “exhibited multiple significant risk factors for self-harm, including, among others, recurring suicidal thoughts and ideations” long before using ChatGPT.

There’s a dark cloud gathering over the company. The case is one of eight lawsuits that have been filed against OpenAI, many of which also allege wrongful death.

Despite arguing in a Tuesday blog post that OpenAI is hoping to handle ongoing litigation with “care, transparency, and respect,” OpenAI’s aggressive legal strategy against Raine’s family strikes other attorneys as unwise.

“As a corporate lawyer one of your jobs is to know when you can make a legal claim but shouldn’t because of the bad public reaction,” lawyer Emory Parker wrote in a Bluesky post. “Like when Disney tried to say that guy couldn’t sue over his wife’s death because of the fine print in a Disney+ trial he signed up for years earlier.”

More on OpenAI: ChatGPT’s Dark Side Encouraged Wave of Suicides, Grieving Families Say

The post OpenAI Says Boy’s Death Was His Own Fault for Using ChatGPT Wrong appeared first on Futurism.

🔗 Sumber: futurism.com