This paper presents a multimodal semantic analysis of accessible Brazilian short films using a frame-based annotation approach. We introduce a subset of the Audition dataset, comprising six short films from the animation and documentary genres. We analysed three communicative modes: original audio, audio description, and visual content. Trained annotators semantically annotated each mode following the FrameNet Brazil multimodal methodology. To compare meaning across modalities, we used cosine similarity over frame-semantic representations. Results show that audio description aligns more closely with video content than original audio, reflecting its role in translating visual meaning into language. Our findings demonstrate the effectiveness of frame semantics in modelling meaning across modalities and provide quantitative evidence of audio description as a bridge between visual and verbal communication. The dataset and annotation strategies are a valuable resource for research on multimodal representation, semantic similarity, and accessible media.

This paper presents Framed Multi30K (FM30K), a novel frame-based Brazilian Portuguese multimodal-multilingual dataset which i) extends the Multi30K dataset (Elliot et al., 2016) with 158,915 original Brazilian Portuguese descriptions, and 30,104 Brazilian Portuguese translations from original English descriptions; and ii) adds 2,677,613 frame evocation labels to the 158,915 English descriptions and to the ones created for Brazilian Portuguese; (iii) extends the Flickr30k Entities dataset (Plummer et al., 2015) with 190,608 frames and Frame Elements correlations with the existing phrase-to-region correlations.

This paper presents the Frame2 dataset, a multimodal dataset built from a corpus of a Brazilian travel TV show annotated for FrameNet categories for both the text and image communicative modes. Frame2 comprises 230 minutes of video, which are correlated with 2,915 sentences either transcribing the audio spoken during the episodes or the subtitling segments of the show where the host conducts interviews in English. For this first release of the dataset, a total of 11,796 annotation sets for the sentences and 6,841 for the video are included. Each of the former includes a target lexical unit evoking a frame or one or more frame elements. For each video annotation, a bounding box in the image is correlated with a frame, a frame element and lexical unit evoking a frame in FrameNet.