Crowdsourcing a Dataset of Audio Captions

Abstract

Audio captioning is a novel field of multi-modal translation and it is the task of creating a textual description of the content of an audio signal (e.g. "people talking in a big room"). The creation of a dataset for this task requires a considerable amount of work, rendering the crowdsourcing a very attractive option. In this paper we present a three steps based framework for crowdsourcing an audio captioning dataset, based on concepts and practises followed for the creation of widely used image captioning and machine translations datasets. During the first step initial captions are gathered. A grammatically corrected and/or rephrased version of each initial caption is obtained in second step. Finally, the initial and edited captions are rated, keeping the top ones for the produced dataset. We objectively evaluate the impact of our framework during the process of creating an audio captioning dataset, in terms of diversity and amount of typographical errors in the obtained captions. The obtained results show that the resulting dataset has less typographical errors than the initial captions, and on average each sound in the produced dataset has captions with a Jaccard similarity of 0.24, roughly equivalent to two ten-word captions having in common four words with the same root, indicating that the captions are dissimilar while they still contain some of the same information.

Caption Feature Space Regularization for Audio Captioning

“…The audio captioning task is firstly introduced in [1], which proposed the commercial ProSound Effects [6] audio corpus as a proof of concept. The paper proposed a BiGRU [7] based encoder-decoder model to generate audio captions.…”

“…Even for people, precisely distinguishing events in audio can be difficult, let alone effectively describing the contents of given audio, because the description is often dependent on the situation or context as much as the audio itself. Therefore, due to the ambiguity of audio, different persons may have varying perceptions of the same audio, which will result in the semantic disparity of audio captions [2], for example, a thin plastic rattling could be perceived as a fire crackling [6] (as shown in Fig. 1).…”