Time Reversal as Self-Supervision

By Suraj Nair, Mohammad Babaeizadeh, Chelsea Finn, Sergey Levine, Vikash Kumar. 2018

12 citations recorded by the source corpus.

A longstanding challenge in robot learning for manipulation tasks has been the ability to generalize to varying initial conditions, diverse objects, and changing objectives. Learning based approaches have shown promise in producing robust policies, but require heavy supervision to efficiently learn precise control, especially from visual inputs. We propose a novel self-supervision technique that uses time-reversal to learn goals and provide a high level plan to reach them. In particular, we introduce the time-reversal model (TRM), a self-supervised model which explores outward from a set of goal states and learns to predict these trajectories in reverse. This provides a high level plan towards goals, allowing us to learn complex manipulation tasks with no demonstrations or exploration at test time. We test our method on the domain of assembly, specifically the mating of tetris-style block pairs. Using our method operating atop visual model predictive control, we are able to assemble tetris blocks on a physical robot using only uncalibrated RGB camera input, and generalize to unseen block pairs. sites.google.com/view/time-reversal

Subfields: Aerospace Engineering, Control and Systems Engineering, Media Technology

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Aerospace Engineering

1. Sergey Levine
2. Andy Zeng
3. Chelsea Finn
4. Suraj Nair
5. Shuran Song
6. Johnny Lee
7. Sudeep Dasari
8. Pieter Abbeel

In Control and Systems Engineering

1. Chelsea Finn
2. Sergey Levine
3. Suraj Nair
4. Andy Zeng
5. Frederik D. Ebert
6. Sudeep Dasari
7. Pieter Abbeel
8. Shuran Song

In Media Technology

1. Chelsea Finn
2. Suraj Nair
3. Sergey Levine
4. Shuran Song
5. Andy Zeng

Citations in the corpus

• RoboNet: Large-Scale Multi-Robot Learning (2019, methodology): “This method is related to a range of recently proposed techniques that use predictive models for vision-based control [28, 29, 30, 31, 32].”
• Form2Fit: Learning Shape Priors for Generalizable Assembly from Disassembly (2019, methodology): “More recent works [11], [12] eliminate the need for accurate state estimation by learning a policy that directly maps from raw pixel observations to actions through reinforcement learning.”
• Hierarchical Foresight: Self-Supervised Learning of Long-Horizon Tasks via Visual Subgoal Generation (2019, methodology): “…et al., 2018; Lange et al., 2012; OpenAI et al., 2018; Schenck & Fox, 2016; Matas et al., 2018; James et al., 2017; Singh et al., 2019), including in goal-conditioned settings (Kaelbling, 1993; Schaul et al., 2015; Andrychowicz et al., 2017; Sadeghi et al., 2018; Sadeghi, 2019; Nair et al., 2018a).”