DeepAI AI Chat
Log In Sign Up

CASSL: Curriculum Accelerated Self-Supervised Learning

by   Adithyavairavan Murali, et al.
Carnegie Mellon University

Recent self-supervised learning approaches focus on using a few thousand data points to learn policies for high-level, low-dimensional action spaces. However, scaling this framework for high-dimensional control require either scaling up the data collection efforts or using a clever sampling strategy for training. We present a novel approach - Curriculum Accelerated Self-Supervised Learning (CASSL) - to train policies that map visual information to high-level, higher- dimensional action spaces. CASSL orders the sampling of training data based on control dimensions: the learning and sampling are focused on few control parameters before other parameters. The right curriculum for learning is suggested by variance-based global sensitivity analysis of the control space. We apply our CASSL framework to learning how to grasp using an adaptive, underactuated multi-fingered gripper, a challenging system to control. Our experimental results indicate that CASSL provides significant improvement and generalization compared to baseline methods such as staged curriculum learning (8 improvement) tested on a set of novel objects.


page 2

page 3

page 7


Scaling and Benchmarking Self-Supervised Visual Representation Learning

Self-supervised learning aims to learn representations from the data its...

How Severe is Benchmark-Sensitivity in Video Self-Supervised Learning?

Despite the recent success of video self-supervised learning, there is m...

PointSmile: Point Self-supervised Learning via Curriculum Mutual Information

Self-supervised learning is attracting wide attention in point cloud pro...

Exploiting map information for self-supervised learning in motion forecasting

Inspired by recent developments regarding the application of self-superv...

Self-supervised visual feature learning with curriculum

Self-supervised learning techniques have shown their abilities to learn ...

Learning body-affordances to simplify action spaces

Controlling embodied agents with many actuated degrees of freedom is a c...