| 1 |
Cross-Situational Learning Towards Robot Grounding
|
|
|
|
In: https://hal.archives-ouvertes.fr/hal-03628290 ; 2022 (2022)
|
|
Abstract:
How do children acquire language through unsupervised or noisy supervision? How do their brain process language? We take this perspective to machine learning and robotics, where part of the problem is understanding how language models can perform grounded language acquisition through noisy supervision and discussing how they can account for brain learning dynamics. Most prior works have tracked the co-occurrence between single words and referents to model how infants learn wordreferent mappings. This paper studies cross-situational learning (CSL) with full sentences: we want to understand brain mechanisms that enable children to learn mappings between words and their meanings from full sentences in early language learning. We investigate the CSL task on a few training examples with two sequence-based models: (i) Echo State Networks (ESN) and (ii) Long-Short Term Memory Networks (LSTM). Most importantly, we explore several word representations including One-Hot, GloVe, pretrained BERT, and fine-tuned BERT representations (last layer token representations) to perform the CSL task. We apply our approach to three diverse datasets (two grounded language datasets and a robotic dataset) and observe that (1) One-Hot, GloVe, and pretrained BERT representations are less efficient when compared to representations obtained from fine-tuned BERT. (2) ESN online with final learning (FL) yields superior performance over ESN online continual learning (CL), offline learning, and LSTMs, indicating the more biological plausibility of ESNs and the cognitive process of sentence reading. (2) LSTM with fewer hidden units showcases higher performance for small datasets, but LSTM with more hidden units is Cross-Situational Learning needed to perform reasonably well on larger corpora. (4) ESNs demonstrate better generalization than LSTM models for increasingly large vocabularies. Overall, these models are able to learn from scratch to link complex relations between words and their corresponding meaning concepts, handling polysemous and synonymous words. Moreover, we argue that such models can extend to help current human-robot interaction studies on language grounding and better understand children's developmental language acquisition. We make the code publicly available * .
|
|
Keyword:
[INFO.INFO-AI]Computer Science [cs]/Artificial Intelligence [cs.AI]; [INFO.INFO-CL]Computer Science [cs]/Computation and Language [cs.CL]; [INFO.INFO-LG]Computer Science [cs]/Machine Learning [cs.LG]; [INFO.INFO-NE]Computer Science [cs]/Neural and Evolutionary Computing [cs.NE]; [INFO.INFO-RB]Computer Science [cs]/Robotics [cs.RO]; [SDV.NEU]Life Sciences [q-bio]/Neurons and Cognition [q-bio.NC]; BERT; cross-situational learning; echo state networks; grounded language; LSTM
|
|
URL: https://hal.archives-ouvertes.fr/hal-03628290
https://hal.archives-ouvertes.fr/hal-03628290v2/file/Journal_of_Social_and_Robotics.pdf
https://hal.archives-ouvertes.fr/hal-03628290v2/document
|
|
BASE
|
|
Hide details
|
|
|
| 2 |
Cross-Situational Learning Towards Robot Grounding
|
|
|
|
In: https://hal.archives-ouvertes.fr/hal-03628290 ; 2022 (2022)
|
|
BASE
|
|
Show details
|
|
|
| 3 |
Hybrid models of cognition: The influence of modal and amodal cues in language processing tasks
|
|
|
|
BASE
|
|
Show details
|
|
|
| 4 |
Abstract concepts, language and sociality: from acquisition to inner speech ...
|
|
|
|
BASE
|
|
Show details
|
|
|
| 5 |
Quantifying Social Semantics: an Inclusive Definition of Socialness and Ratings for 8,388 English Words ...
|
|
|
|
BASE
|
|
Show details
|
|
|
| 8 |
Dynamic grounding of emotion concepts.
|
|
|
|
In: Philosophical transactions of the Royal Society of London. Series B, Biological sciences, vol 373, iss 1752 (2018)
|
|
BASE
|
|
Show details
|
|
|
| 9 |
The Perceptual Basis of Abstract Concepts in Polysemy Networks – An Interdisciplinary Study
|
|
|
|
In: http://rave.ohiolink.edu/etdc/view?acc_num=case1511400502977642 (2018)
|
|
BASE
|
|
Show details
|
|
|
| 10 |
The semantic content of abstract concepts: a property listing study of 296 abstract words ...
|
|
|
|
BASE
|
|
Show details
|
|
|
| 11 |
L'Embodied Cognition e la sfida dei concetti astratti. Un approccio multidimensionale
|
|
|
|
In: Rivista internazionale di Filosofia e Psicologia; V. 9, N. 3 (2018); 239-253 ; 2239-2629 ; 2039-4667 (2018)
|
|
BASE
|
|
Show details
|
|
|
| 12 |
Conceptual semantics as grounded in personal experience ...
|
|
|
|
BASE
|
|
Show details
|
|
|
| 14 |
Abstract Concepts and Pictures of Real-World Situations Activate One Another.
|
|
|
|
In: Psychology Publications (2018)
|
|
BASE
|
|
Show details
|
|
|
| 15 |
Meaning Transfer in a Cobranding Context: The Role of the Sensory Signature
|
|
|
|
BASE
|
|
Show details
|
|
|
| 16 |
The Peculiarity of Emotional Words: A Grounded Approach
|
|
|
|
In: Rivista internazionale di Filosofia e Psicologia; V. 8, N. 2 (2017); 124-133 ; 2239-2629 ; 2039-4667 (2017)
|
|
BASE
|
|
Show details
|
|
|
| 17 |
Abstract concepts and aging: An embodied and grounded perspective
|
|
|
|
BASE
|
|
Show details
|
|
|
| 18 |
Differential Impairments of Upper and Lower Limb Movements Influence Action Verb Processing in Parkinson Disease
|
|
|
|
In: Psychology Publications (2017)
|
|
BASE
|
|
Show details
|
|
|
| 19 |
Neural Adaptation Effects in Conceptual Processing
|
|
|
|
In: Behavioral Sciences ; Volume 5 ; Issue 3 ; Pages 353-371 (2015)
|
|
BASE
|
|
Show details
|
|
|
| 20 |
Grounded narrative inquiry into language teacher cognition: Stories and case studies on English language teaching in South Korea
|
|
|
|
BASE
|
|
Show details
|
|
|
|
|
