Robot Manipulation Generalization
18 papers with code • 2 benchmarks • 2 datasets
Benchmarks
These leaderboards are used to track progress in Robot Manipulation Generalization
| Trend | Dataset | Best Model | Paper | Code | Compare |
|---|---|---|---|---|---|
|
|||||
|
RLBench
The COLOSSEUM
Most implemented papers
Segment Anything
facebookresearch/segment-anything
•
•
ICCV 2023
We introduce the Segment Anything (SA) project: a new task, model, and dataset for image segmentation.
SAM 2: Segment Anything in Images and Videos
We present Segment Anything Model 2 (SAM 2), a foundation model towards solving promptable visual segmentation in images and videos.
Instruction-driven history-aware policies for robotic manipulations
In human environments, robots are expected to accomplish a variety of manipulation tasks given simple natural language instructions.
Segment Anything for Videos: A Systematic Survey
liliu-avril/Awesome-Segment-Anything
•
•
To address this gap, this work conducts a systematic review on SAM for videos in the era of foundation models.
Sam2Rad: A Segmentation Model for Medical Images with Learnable Prompts
SAM and its variants often fail to segment structures in ultrasound (US) images due to domain shift.
Masked Visual Pre-training for Motor Control
ir413/mvp
•
•
This paper shows that self-supervised visual pre-training from real-world images is effective for learning motor control tasks from pixels.
R3M: A Universal Visual Representation for Robot Manipulation
We study how visual representations pre-trained on diverse human video data can enable data-efficient learning of downstream robotic manipulation tasks.
Perceiver-Actor: A Multi-Task Transformer for Robotic Manipulation
peract/peract
•
•
With this formulation, we train a single multi-task Transformer for 18 RLBench tasks (with 249 variations) and 7 real-world tasks (with 18 variations) from just a few demonstrations per task.
RVT: Robotic View Transformer for 3D Object Manipulation
NVlabs/RVT
•
•
In simulations, we find that a single RVT model works well across 18 RLBench tasks with 249 task variations, achieving 26% higher relative success than the existing state-of-the-art method (PerAct).
PolarNet: 3D Point Clouds for Language-Guided Robotic Manipulation
The ability for robots to comprehend and execute manipulation tasks based on natural language instructions is a long-term goal in robotics.
