CIDEr swMATH ID: 26599 Software Authors: Ramakrishna Vedantam, C. Lawrence Zitnick, Devi Parikh Description: CIDEr: Consensus-based Image Description Evaluation. Automatically describing an image with a sentence is a long-standing challenge in computer vision and natural language processing. Due to recent progress in object detection, attribute classification, action recognition, etc., there is renewed interest in this area. However, evaluating the quality of descriptions has proven to be challenging. We propose a novel paradigm for evaluating image descriptions that uses human consensus. This paradigm consists of three main parts: a new triplet-based method of collecting human annotations to measure consensus, a new automated metric (CIDEr) that captures consensus, and two new datasets: PASCAL-50S and ABSTRACT-50S that contain 50 sentences describing each image. Our simple metric captures human judgment of consensus better than existing metrics across sentences generated by various sources. We also evaluate five state-of-the-art image description approaches using this new protocol and provide a benchmark for future comparisons. A version of CIDEr named CIDEr-D is available as a part of MS COCO evaluation server to enable systematic evaluation and benchmarking. Homepage: https://arxiv.org/abs/1411.5726 Related Software: BLEU; Rouge; Flickr30K; EfficientDet; YOLO; ImageNet; VQA; MS-COCO; BERT; DenseCap; Adam; Caffe; ConceptNet; BERTScore; Tensor2Tensor; AnchorNet; Soft scissors; MVSNet; NIMA; EfficientNet Cited in: 4 Documents all top 5 Cited by 18 Authors 1 Baldwin, Timothy 1 Belgodere, Brian 1 Dognin, Pierre 1 Huang, Feicheng 1 Koto, Fajri 1 Lau, Jey Han 1 Li, Zhixin 1 Ma, Huifang 1 Melnyk, Igor 1 Mroueh, Youssef 1 Padhi, Inkit 1 Rigotti, Mattia 1 Ross, Jarret 1 Schiff, Yair 1 Szeliski, Richard 1 Wei, Haiyang 1 Young, Richard A. 1 Zhang, Canlong Cited in 3 Serials 2 The Journal of Artificial Intelligence Research (JAIR) 1 Machine Learning 1 Texts in Computer Science Cited in 1 Field 4 Computer science (68-XX) Citations by Year