OpenVLA
A large-scale open-source vision-language-action (VLA) model developed by the academic community. It follows a monolithic model architecture that directly generates robot motion commands from visual and language inputs.
Core Facts
VLA models integrate visual perception, language understanding, and robot action planning into a unified model, typically using a pretrained large vision-language model as the backbone, with representative works including Google's RT-2 and OpenVLA.
RT-2 first demonstrated the feasibility of the VLA architecture by discretizing robot actions into tokens and having the vision-language model directly output action sequences.
Timeline (last 90 days)
Representative VLA works include RT-2 and OpenVLA, which are typically pretrained on massive image-text-action triplet data and possess a certain degree of open-world generalization capability.
Representative works of VLA models include RT-2, OpenVLA, and others, with parameter scales ranging from several billion to several tens of billions.
Typical representatives of VLA models include RT-2, OpenVLA and GR00T, which are usually pretrained jointly on internet-scale image-text data and large-scale robotic manipulation data.
Typical representatives of action models are vision-language-action models (VLA), such as RT-2 and OpenVLA, which usually use a pretrained multimodal large model as the backbone and, after fine-tuning on robot data, directly output joint positions, end-effector trajectories or discrete action tokens.
In the field of robot foundation models, NVIDIA's GR00T series already exists, and in the open-source community, VLA models such as RT-2 and OpenVLA are iterating rapidly.
OpenVLA uses an open-source vision-language model with about 7 billion parameters as its backbone and is fine-tuned on the Open X-Embodiment dataset, with downloadable weights provided.
Representative VLA works include RT-2, OpenVLA, π0, etc.
Representative works in VLA models include Google's RT-2, OpenVLA, and π0.
RT-2, PaLM-E, and OpenVLA are robot control frameworks based on pretrained models.
VLA models integrate visual perception, language understanding, and robot action planning into a unified model, typically using a pretrained large vision-language model as the backbone, with representative works including Google's RT-2 and OpenVLA.
4 more timeline events
All Facts (14)
VLA models integrate visual perception, language understanding, and robot action planning into a unified model, typically using a pretrained large vision-language model as the backbone, with representative works including Google's RT-2 and OpenVLA.
75%VerifiedRT-2 first demonstrated the feasibility of the VLA architecture by discretizing robot actions into tokens and having the vision-language model directly output action sequences.
65%UnverifiedRepresentative VLA works include RT-2 and OpenVLA, which are typically pretrained on massive image-text-action triplet data and possess a certain degree of open-world generalization capability.
50%UnverifiedRepresentative works of VLA models include RT-2, OpenVLA, and others, with parameter scales ranging from several billion to several tens of billions.
50%UnverifiedTypical representatives of VLA models include RT-2, OpenVLA and GR00T, which are usually pretrained jointly on internet-scale image-text data and large-scale robotic manipulation data.
50%UnverifiedTypical representatives of action models are vision-language-action models (VLA), such as RT-2 and OpenVLA, which usually use a pretrained multimodal large model as the backbone and, after fine-tuning on robot data, directly output joint positions, end-effector trajectories or discrete action tokens.
50%UnverifiedOpenVLA uses an open-source vision-language model with about 7 billion parameters as its backbone and is fine-tuned on the Open X-Embodiment dataset, with downloadable weights provided.
50%UnverifiedIn the field of robot foundation models, NVIDIA's GR00T series already exists, and in the open-source community, VLA models such as RT-2 and OpenVLA are iterating rapidly.
50%UnverifiedRepresentative VLA works include RT-2, OpenVLA, π0, etc.
50%UnverifiedRepresentative works in VLA models include Google's RT-2, OpenVLA, and π0.
50%UnverifiedRT-2, PaLM-E, and OpenVLA are robot control frameworks based on pretrained models.
50%UnverifiedOpenVLA, π0, RT-2, and RDT-1B all belong to the "monolithic" VLA model category.
50%UnverifiedVision-Language-Action models (VLAs) unify visual perception, language understanding, and action generation within a single framework, enabling robots to receive natural language instructions and directly output control signals
50%UnverifiedRT-2是Google DeepMind的Robotics Transformer 2，与OpenVLA等大规模机器人基础模型需要覆盖多样化场景、包含精确空间信息和语义标注的真实世界数据
50%