Evaluating Automatically Generated Phoneme Captions for Images
This work addresses the problem of evaluating image-to-speech systems for researchers, but it is incremental as it builds on existing tasks and metrics.
The paper tackled the evaluation of Image2Speech systems by implementing a system that generates phoneme captions, which outperformed the original on the Flickr8k corpus, and found that BLEU4 had the highest correlation with human ratings among tested metrics.
Image2Speech is the relatively new task of generating a spoken description of an image. This paper presents an investigation into the evaluation of this task. For this, first an Image2Speech system was implemented which generates image captions consisting of phoneme sequences. This system outperformed the original Image2Speech system on the Flickr8k corpus. Subsequently, these phoneme captions were converted into sentences of words. The captions were rated by human evaluators for their goodness of describing the image. Finally, several objective metric scores of the results were correlated with these human ratings. Although BLEU4 does not perfectly correlate with human ratings, it obtained the highest correlation among the investigated metrics, and is the best currently existing metric for the Image2Speech task. Current metrics are limited by the fact that they assume their input to be words. A more appropriate metric for the Image2Speech task should assume its input to be parts of words, i.e. phonemes, instead.