Changsong Liu

About

Changsong Liu is a pioneering researcher in human-robot interaction, with a particular focus on situated dialogue, referential grounding, and collaborative language communication. His work addresses one of the most fundamental challenges in robotics: bridging the perceptual gap between humans and robots who share a physical environment yet possess radically different capabilities for perceiving and representing it. Liu's most influential contributions center on developing computational frameworks that allow robots to establish common ground with human partners through dialogue. His highly cited 2014 paper on collaborative human-robot dialogue (75 citations) and his 2016 work on collaborative language grounding (67 citations) introduced innovative approaches to aligning mismatched perceptual representations, enabling more natural and effective communication. His graph-based representations and probabilistic labeling techniques offer elegant solutions to the complex problem of referential grounding in dynamic environments. Beyond dialogue, Liu has made significant strides in machine learning for robotics, developing methods for robots to acquire task knowledge jointly from natural language instruction and visual demonstration using hierarchical And-Or Graph representations. Collectively accumulating nearly 300 citations, his body of work has meaningfully advanced the field of cognitively capable, communicatively competent robots, making him an essential voice in human-robot collaboration research.

Research Focus

Key Achievements

Top Papers

1. 1
   Collaborative effort towards common ground in situated human-robot dialogue

   75 citations · 2014
2. 2
   Collaborative Language Grounding Toward Situated Human‐Robot Dialogue

   67 citations · 2016
3. 3
   Towards mediating shared perceptual basis in situated dialogue

   42 citations · 2012
4. 4
   Jointly Learning Grounded Task Structures from Language Instruction and Visual Demonstration

   42 citations · 2016
5. 5
   Probabilistic Labeling for Efficient Referential Grounding based on Collaborative Discourse

   16 citations · 2014
6. 6
   Ambiguities in Spatial Language Understanding in Situated Human Robot Dialogue

   16 citations · 2010
7. 7
   Learning to Mediate Perceptual Differences in Situated Human-Robot Dialogue

   15 citations · 2015
8. 8
   Integrating word acquisition and referential grounding towards physical world interaction

   10 citations · 2012
9. 9
   Task Learning through Visual Demonstration and Situated Dialogue.

   7 citations · 2016

Key Collaborators

Contact & Links