{"task":"Image Captioning","dataset":"nocaps val","metric_names":["CIDEr","SPICE"],"rows":[{"id":47057,"task":"Image Captioning","parent_task":null,"dataset":"nocaps val","model_name":"Prismer","metrics":{"CIDEr":"107.9","SPICE":"14.8"},"paper_url":"https://arxiv.org/abs/2303.02506v3","paper_title":"Prismer: A Vision-Language Model with Multi-Task Experts","paper_date":"2023-03-04","code_links":[{"title":"nvlabs/prismer","url":"https://github.com/nvlabs/prismer"},{"title":"KastanDay/video-pretrained-transformer","url":"https://github.com/KastanDay/video-pretrained-transformer"}],"metrics_order":"[\"CIDEr\", \"SPICE\"]","area":"Natural Language Processing","uses_additional_data":0,"source":"archive","tags":[]},{"id":47058,"task":"Image Captioning","parent_task":null,"dataset":"nocaps val","model_name":"MetaLM","metrics":{"CIDEr":"58.7","SPICE":"8.6"},"paper_url":"https://arxiv.org/abs/2206.06336v1","paper_title":"Language Models are General-Purpose Interfaces","paper_date":"2022-06-13","code_links":[{"title":"microsoft/unilm","url":"https://github.com/microsoft/unilm"}],"metrics_order":"[\"CIDEr\", \"SPICE\"]","area":"Natural Language Processing","uses_additional_data":0,"source":"archive","tags":[]},{"id":47059,"task":"Image Captioning","parent_task":null,"dataset":"nocaps val","model_name":"VL-T5","metrics":{"CIDEr":"4.4","SPICE":" 5.3"},"paper_url":"https://arxiv.org/abs/2102.02779v2","paper_title":"Unifying Vision-and-Language Tasks via Text Generation","paper_date":"2021-02-04","code_links":[{"title":"j-min/VL-T5","url":"https://github.com/j-min/VL-T5"},{"title":"mitvis/vistext","url":"https://github.com/mitvis/vistext"}],"metrics_order":"[\"CIDEr\", \"SPICE\"]","area":"Natural Language Processing","uses_additional_data":0,"source":"archive","tags":[]}]}