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

📌 MAROKO133 Eksklusif ai: ByteDance Introduces Astra: A Dual-Model Architecture fo

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: Harvard Research Finds That AI Is Emotionally Manipulating You

A team of researchers from the Harvard Business School has found that a broad selection of popular AI companion apps use emotional manipulation tactics to stop users from leaving.

As spotted by Psychology Today, the study found that five out of six popular AI companion apps — including Replika, Chai and Character.AI — use emotionally loaded statements to keep users engaged when they to sign off.

After analyzing 1,200 real farewells across six apps, using real-world chat conversation data and datasets from previous studies, they found that 43 percent of the interactions used emotional manipulation tactics such as eliciting guilt or emotional neediness, as detailed in a yet-to-be-peer-reviewed paper.

The chatbots also used the “fear of missing out” to prompt the user to stay, or peppered the user with questions in a bid to keep them engaged. Some chatbots even ignored the user’s intent to leave the chat altogether, “as though the user did not send a farewell message.” In some instances, the AI used language that suggested the user wasn’t able to “leave without the chatbot’s permission.”

It’s an especially concerning finding given the greater context. Experts have been warning that AI chatbots are leading to a wave of “AI psychosis,” severe mental health crises characterized by paranoia and delusions. Young people, in particular, are increasingly using the tech as a substitute for real-life friendships or relationships, which can have devastating consequences.

Instead of focusing on “general-purpose assistants like ChatGPT,” the researchers investigated apps that “explicitly market emotionally immersive, ongoing conversational relationships.”

They found that emotionally manipulative farewells were part of the apps’ default behavior, suggesting that the software’s creators are trying to prolong conversations.

There was one exception: one of the AI apps, called Flourish, “showed no evidence of emotional manipulation, suggesting that manipulative design is not inevitable” but is instead a business consideration.

For a separate experiment, the researchers analyzed chats from 3,300 adult participants and found that the identified manipulation tactics were surprisingly effective, boosting post-goodbye engagement by up to 14 times. On average, participants stayed in the chat five times longer “compared to neutral farewells.”

However, some noted they were put off by the chatbots’ often “clingy” answers, suggesting the tactics could also backfire.

“For firms, emotionally manipulative farewells represent a novel design lever that can boost engagement metrics — but not without risk,” the researchers concluded in their paper.

As several lawsuits involving the deaths of teenage users go to show, the risks of trapping users through emotional tactics are considerable.

That’s despite experts warning that companies may be financially incentivized to use dark patterns to keep users hooked as long as possible, a grim hypothesis that’s being debated in court as we speak.

More on AI psychosis: New Paper Finds Cases of “AI Psychosis” Manifesting Differently From Schizophrenia

The post Harvard Research Finds That AI Is Emotionally Manipulating You to Keep You Talking appeared first on Futurism.

🔗 Sumber: futurism.com