A morphable template framework for robot learning by demonstration: Integrating one-shot and incremental learning approaches

There are several deep learning approaches that can be applied for analyzing situations in real world problems and inventing their solution in a scientific technique. Supervised data mining methods that predicts instance values, using previously obtained results from already collected data are pretty popular due to their intelligence in machine learning area. Stream data is continuous form of data which can be handled by using incremental learning approach. Stream data learning may face several challenges in real world like concept drift or class imbalance. Concept drift occurs in non-stationary environment where data distribution generation function is dynamic in nature and has no fixed formula to predict the future data distribution nature. Neural network techniques are intelligent enough to improve performance of algorithmic systems that work in such problem domains. This chapter briefly describes how MLP technique is integrated in system so that the system becomes a complete framework for handling unbalanced data with concept drift in the incremental learning strategies.

Download Full-text

Bayesian approach to incremental batch learning on forest cover sensor data for multiclass classification

Concurrent Engineering ◽

10.1177/1063293x211058450 ◽

2021 ◽

pp. 1063293X2110584

Author(s):

Venkata Vara Prasad D ◽

Lokeswari Y Venkataramana ◽

Saraswathi S ◽

Sarah Mathew ◽

Snigdha V

Keyword(s):

Bayesian Approach ◽

Incremental Learning ◽

Bayesian Model ◽

Forest Cover ◽

Bayesian Framework ◽

Training Data ◽

Sensor Data ◽

Learning Approaches ◽

New Class ◽

Multiple Levels

Deep neural networks can be used to perform nonlinear operations at multiple levels, such as a neural network that is composed of many hidden layers. Although deep learning approaches show good results, they have a drawback called catastrophic forgetting, which is a reduction in performance when a new class is added. Incremental learning is a learning method where existing knowledge should be retained even when new data is acquired. It involves learning with multiple batches of training data and the newer learning sessions do not require the data used in the previous iterations. The Bayesian approach to incremental learning uses the concept of the probability distribution of weights. The key idea of Bayes theorem is to find an updated distribution of weights and biases. In the Bayesian framework, the beliefs can be updated iteratively as the new data comes in. Bayesian framework allows to update the beliefs iteratively in real-time as data comes in. The Bayesian model for incremental learning showed an accuracy of 82%. The execution time for the Bayesian model was lesser on GPU (670 s) when compared to CPU (1165 s).

Download Full-text

A nonparametric Bayesian approach toward robot learning by demonstration

Robotics and Autonomous Systems ◽

10.1016/j.robot.2012.02.005 ◽

2012 ◽

Vol 60 (6) ◽

pp. 789-802 ◽

Cited By ~ 19

Author(s):

Sotirios P. Chatzis ◽

Dimitrios Korkinof ◽

Yiannis Demiris

Keyword(s):

Bayesian Approach ◽

Robot Learning ◽

Nonparametric Bayesian ◽

Learning By Demonstration

Download Full-text

Invariant signature description and trajectory reproduction for robot Learning by Demonstration

2008 IEEE/RSJ International Conference on Intelligent Robots and Systems ◽

10.1109/iros.2008.4650646 ◽

2008 ◽

Cited By ~ 2

Author(s):

Shandong Wu ◽

Y.F. Li ◽

Jianwei Zhang

Keyword(s):

Robot Learning ◽

Learning By Demonstration

Download Full-text

Object Detection-Based One-Shot Imitation Learning with an RGB-D Camera

Applied Sciences ◽

10.3390/app10030803 ◽

2020 ◽

Vol 10 (3) ◽

pp. 803 ◽

Cited By ~ 2

Author(s):

Quanquan Shao ◽

Jin Qi ◽

Jin Ma ◽

Yi Fang ◽

Weiming Wang ◽

...

Keyword(s):

Object Detection ◽

Visual Information ◽

Depth Map ◽

Robot Learning ◽

Target Object ◽

Fine Tuning ◽

Great Success ◽

Learning Approaches ◽

Policy Network ◽

End To End

End-to-end robot learning has achieved a great success for robots to obtain various manipulation skills. It learns a function which maps visual information to robotic action directly. Because of the diversity of target objects, most end-to-end robot learning approaches have focused on a single object-specific task with a limited capability of generalization. In this work, an object detection-based one-shot learning method is proposed, which separates the semantic understanding from robot control. It enables a robot to acquire similar manipulation skills efficiently and to have the ability to cope with new objects with a single demonstration. This approach mainly has two modules: the object detection network and the motion policy network. With RGB images, the object detection network tries to output the task-related semantic keypoint of the target object, which is the center of the container in this application, and the motion policy network generates the motion action based on the depth map and the detected keypoint. To evaluate this proposed pipeline, a series of experiments are conducted on typical placing tasks in different simulation scenarios and, additionally, the learned policy is transferred from simulation to the real world without any fine-tuning.

Download Full-text

Clustering versus Incremental Learning Multi-Codebook Fuzzy Neural Network for Multi-Modal Data Classification

Computation ◽

10.3390/computation8010006 ◽

2020 ◽

Vol 8 (1) ◽

pp. 6

Author(s):

Muhammad Anwar Ma’sum ◽

Hadaiq Rolis Sanabila ◽

Petrus Mursanto ◽

Wisnu Jatmiko

Keyword(s):

Neural Network ◽

Incremental Learning ◽

Fuzzy Neural Network ◽

Data Classification ◽

Benchmark Dataset ◽

Learning Approaches ◽

Clustering Methods ◽

Modal Data ◽

Fuzzy Neural ◽

Benchmark Datasets

One of the challenges in machine learning is a classification in multi-modal data. The problem needs a customized method as the data has a feature that spreads in several areas. This study proposed a multi-codebook fuzzy neural network classifiers using clustering and incremental learning approaches to deal with multi-modal data classification. The clustering methods used are K-Means and GMM clustering. Experiment result, on a synthetic dataset, the proposed method achieved the highest performance with 84.76% accuracy. Whereas on the benchmark dataset, the proposed method has the highest performance with 79.94% accuracy. The proposed method has 24.9% and 4.7% improvements in synthetic and benchmark datasets respectively compared to the original version. The proposed classifier has better accuracy compared to a popular neural network with 10% and 4.7% margin in synthetic and benchmark dataset respectively.

Download Full-text