The fast-paced development of learning models and training tactics in the field of artificial intelligence is motivating robotics community to employ machine learning-based data-driven approaches, and noticeable achievements have recently been reported in fundamental robotics fields like mapping and navigation, grasping, and even dexterous in-hand manipulations. In social robotics, some pioneering works have also been appearing which attempt to tackle the challenges of providing robots with social skills, such as non-verbal interaction behaviors and co-speech gestures, using deep learning. 

As exemplified by these early works, applying machine learning to social robotics poses several technical issues. First, there are just a few small datasets for training social interactions, that inhibits active adoption of learning-based technologies to social robotics. Building datasets for social interaction is very difficult as it usually involves annotation of variety of multimodal social signals at the 10-1 seconds scale. We need to invent novel methods and systems to efficiently collect training data for social interaction. Second, learning interactions involves complex cognitive strategies, e.g. combinations of imitation, reinforcement and personalization, that cannot simply be realized with a typical single learning algorithm. It is important to come up with effective learning models and tactics especially designed for training social interactions. Third, there is an innate ambiguity in evaluating goodness of learned models in social robotics. How can you define a loss function for optimizing social behaviors? Searching for proper strategies to evaluate the quality of social interactions is essential for machine learning technologies in social robotics. 

With this special issue, we intend to collect research papers that share recent results of machine learning-based social robotics and discuss about the aforementioned issues with regard to machine learning for social robotics.

We are inviting research papers at the intersection of machine learning and social robotics, including but not limited to:

  • Datasets for training social intelligence
  • Learning-based approaches for social human-robot interaction
  • Social perception and context awareness
  • Affect and emotion recognition
  • Social behavior learning and generation
  • Social/Emotional expression generation
  • Social interaction modelling and management
  • Social grasping and navigation skills
  • Emotion recognition and expression
  • Social dialog understanding and generation
  • Personalization of social behaviors
  • Learning tactics for social interaction: active learning, tutoring, imitation etc.
  • Learning-based system integration for social robots
  • User studies with learning-based social robots
  • Domain-specific applications of learning-based social interaction: healthcare, receptionist, education etc.

Submission Instructions

Authors should prepare their papers based on template provided by Springer (

[Important] When you submit your paper, please select article type as “Original research” at the beginning. Then in the 4th step “Additional Information“, you should answer “Yes” to the second question “Does this manuscript belong to a special issue?“, then you will have following question to select the special issue. Please select “S.I.: Machine Learning for Social Human-Robot Interaction“.

  1. Visit, and click “Submit manuscript”.
  2. Login to the Editorial Manager as Author.
  3. Click “Submit New Manuscript”.
  4. Select article type as “Original research” at the beginning.
  5. In the 4th step “Additional Information”, you should answer “Yes” to the second question “Does this manuscript belong to a special issue?”
  6. Then you will have following question to select the special issue. Please select “S.I.: Machine Learning for Social Human-Robot Interaction”.

Important Dates

  • Paper submission: August 1, 2021
  • Final notification of acceptance: November 30, 2021
  • Online publication: March, 2022

Guest Editors

  • Minsu Jang, ETRI, Korea (minsu[at]
  • Ho Seok Ahn, University of Auckland, New Zealand (hs.ahn[at]
  • Jongsuk Choi, KIST, Korea (pristine70[at]