{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/salsa-a-novel-dataset-for-multimodal-group","title":"SALSA: A Novel Dataset for Multimodal Group Behavior Analysis","arxiv_id":"1506.06882","date":"2015-06-23","proceeding":null,"authors":["Xavier Alameda-Pineda","Jacopo Staiano","Ramanathan Subramanian","Ligia Batrinca","Elisa Ricci","Bruno Lepri","Oswald Lanz","Nicu Sebe"],"abstract":"Studying free-standing conversational groups (FCGs) in unstructured social\nsettings (e.g., cocktail party ) is gratifying due to the wealth of information\navailable at the group (mining social networks) and individual (recognizing\nnative behavioral and personality traits) levels. However, analyzing social\nscenes involving FCGs is also highly challenging due to the difficulty in\nextracting behavioral cues such as target locations, their speaking activity\nand head/body pose due to crowdedness and presence of extreme occlusions. To\nthis end, we propose SALSA, a novel dataset facilitating multimodal and\nSynergetic sociAL Scene Analysis, and make two main contributions to research\non automated social interaction analysis: (1) SALSA records social interactions\namong 18 participants in a natural, indoor environment for over 60 minutes,\nunder the poster presentation and cocktail party contexts presenting\ndifficulties in the form of low-resolution images, lighting variations,\nnumerous occlusions, reverberations and interfering sound sources; (2) To\nalleviate these problems we facilitate multimodal analysis by recording the\nsocial interplay using four static surveillance cameras and sociometric badges\nworn by each participant, comprising the microphone, accelerometer, bluetooth\nand infrared sensors. In addition to raw data, we also provide annotations\nconcerning individuals' personality as well as their position, head, body\norientation and F-formation information over the entire event duration. Through\nextensive experiments with state-of-the-art approaches, we show (a) the\nlimitations of current methods and (b) how the recorded multiple cues\nsynergetically aid automatic analysis of social interactions. SALSA is\navailable at http://tev.fbk.eu/salsa.","url_abs":"http://arxiv.org/abs/1506.06882v1","url_pdf":"http://arxiv.org/pdf/1506.06882v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[],"tasks":[],"methods":[],"datasets_introduced":[{"slug":"salsa","name":"SALSA","full_name":""}],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1506.06882","atlas_url":"https://app.syntology.ai/?focus=1506.06882","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}