Computational Generation of Referring Expressions: A Survey

By Emiel Krahmer, Kees van Deemter. International Conference on Computational Logic, 2012

371 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/ee4fe11884eca58a76ea797f102bbd5c3c67137b

How researchers see this paper

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

In Artificial Intelligence

Citations in the corpus

Differentiable Scene Graphs (2019, background): “Referring Relationships: Several recent studies looked into the task of detecting an entity based on a referring expression [17, 19], while taking context into account.”
ReferItGame: Referring to Objects in Photographs of Natural Scenes (2014, background): “A good survey of work in this area is provided in Krahmer and van Deemter (2012).”
GuessWhat?! Visual Object Discovery through Multi-modal Dialogue (2016, background): “Whenever such expressions unambiguously point to one object, we speak of a referring expression [21].”
Generation and Comprehension of Unambiguous Object Descriptions (2015, methodology): “Referring expression generation is a classic NLP problem (see e.g., [54, 31]).”
Typicality and Object Reference (2013, background): “This is equally true for psycholinguistic 1http://m-mitchell.com/corpora/typicality corpus/ work (Arnold, 2008) and for work on computational models of reference (Dale & Reiter, 1995; Krahmer, van Erk, & Verleg, 2003; Krahmer & van Deemter, 2012).”
The red one!: On learning to refer to things based on discriminative properties (2016, background): “Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics, pages 213–218, Berlin, Germany, August 7-12, 2016. c©2016 Association for Computational Linguistics”
Neural data-to-text generation: A comparison between pipeline and end-to-end architectures (2019, background): “Referring Expression Generation (REG) is the pipeline task responsible for generating the references to the entities of the discourse (Krahmer and van Deemter, 2012).”
See No Evil, Say No Evil: Description Generation from Densely Labeled Images (2014, background): “Visual re-ferring expression generation systems (Krahmer and Van Deemter, 2012; Mitchell et al., 2013; FitzGerald et al., 2013) aim to identify speciﬁc objects, a sub-problem we deal with when describing images more generally.”
Modeling Context Between Objects for Referring Expression Understanding (2016, neutral): “In a retrieval task that focuses on a particular object in an image, the query is called a referring expression [2,3].”
Survey of the State of the Art in Natural Language Generation: Core tasks, applications and evaluation (2017, methodology): “In a human evaluation, Midge was shown to outperform both the system by Kulkarni et al. (2011) and Yang et al. (2011) on a number of criteria, including humanlikeness and correctness.”
Generating Expressions that Refer to Visible Objects (2013, background): “In contrast to most algorithms for the generation of referring expressions, which have aimed to produce distinguishing descriptions when these exist (Krahmer and van Deemter, 2012), the core idea behind this algorithm is to generate what is likely for a speaker in a visual domain.”
Asking for Help Using Inverse Semantics (2014, background): “A key aim of current research is reducing the incidence of these types of failures but eliminating them completely remains an elusive goal.”
Learning to Assemble Neural Module Tree Networks for Visual Grounding (2018, background): “It is one of the core AI tasks for testing the machine comprehension of visual scene and language [18].”
Towards Ecologically Valid Research on Language User Interfaces (2020, neutral): “Generating synthetic data containing a wide variety of such references has been studied but remains challenging (Krahmer and van Deemter, 2012).”
The Abduction of Sherlock Holmes: A Dataset for Visual Abductive Reasoning (2022, methodology): “Following prior work on literal referring expression localization [25, 27, 73] (inter alia), we experiment in two settings: 1) we are given all the groundtruth bounding boxes for an image, and 2) we are given only automatic bounding box proposals from an object detection model.”
Underspecification in Scene Description-to-Depiction Tasks (2022, background): “Within NLP, underspeciﬁcation in descriptions has perhaps been discussed most often in the context of generating referring expressions for objects (see Krahmer and Van Deemter (2012) for a survey).”