This paper focuses on answering fill-in-the-blank style multiple choice questions from the Visual Madlibs dataset. Previous approaches to Visual Question Answering (VQA) have mainly used generic image features from networks trained on the ImageNet dataset, despite the wide scope of questions. In contrast, our approach employs features derived from networks trained for specialized tasks of scene classification, person activity prediction, and person and object attribute prediction. We also present a method for selecting sub-regions of an image that are relevant for evaluating the appropriateness of a putative answer. Visual features are computed both from the whole image and from local regions, while sentences are mapped to a common space using a simple normalized canonical correlation analysis (CCA) model. Our results show a significant improvement over the previous state of the art, and indicate that answering different question types benefits from examining a variety of image cues and carefully choosing informative image sub-regions.
Combining Multiple Cues for Visual Madlibs Question Answering
Dept. of Computer Control and Management Engineering University of Rome, La Sapienza, Italy（罗马大学 Sapienza 计算机控制与管理工程系）；University of Illinois at Urbana Champaign, IL, USA（伊利诺伊大学厄巴纳 - 香槟分校）；University of North Carolina at Chapel Hill, NC, USA（北卡罗来纳大学教堂山分校）；University of Rome, La Sapienza, Italy（罗马大学 Sapienza）
VFaith: Do Large Multimodal Models Really Reason on Seen Images Rather than Previous Memories?
Tsinghua University（清华大学）；Foundation Model Research Center, Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所基础模型研究中心）；School of Artificial Intelligence, University of Chinese Academy of Sciences（中国科学院大学人工智能学院）；ByteDance China（字节跳动中国）；Peng Cheng Laboratory（鹏城实验室）；Wuhan AI Research（武汉人工智能研究）
Remodeling Semantic Relationships in Vision-Language Fine-Tuning
Hangzhou International Innovation Institute, Beihang University（北京航空航天大学杭州国际创新研究院）；School of Artificial Intelligence, Beihang University（北京航空航天大学人工智能学院）；Nanyang Technological University（南洋理工大学）；University of Leicester（莱斯特大学）；School of Astronautics, Beihang University（北京航空航天大学航天学院）
Self-Supervised Moving Vehicle Detection from Audio-Visual Cues
University of Freiburg（弗莱堡大学）；University of Technology Nuremberg（纽伦堡工业大学）
运动车辆的鲁棒检测对于任何自主操作的户外机器人或自动驾驶车辆来说都是一项关键任务。解决这一任务的大多数现代方法都依赖于使用nuScenes或Waymo Open Dataset等大规模车辆检测数据集来训练基于图像的检测器。提供手动注释是一项昂贵且费力的工作，并且在实践中无法很好地扩展。为了解决这个问题，我们提出了一种自我监督的方法，利用视听线索来检测视频中的移动车辆。我们的方法采用对比学习定位车辆的图像，从相应的图像和录音对。在使用真实世界数据集进行的广泛实验中，我们证明了我们的方法可以准确检测移动车辆，并且不需要手动注释。我们还表明，我们的模型可以作为一个老师来监督一个只检测音频的模型。该学生模型对光照变化不敏感，因此有效地弥补了仅利用视觉作为主要模态的模型所固有的领域差距。
Transductive Zero-Shot Learning by Decoupled Feature Generation
Istituto Italiano di Tecnologia（意大利技术研究院）；University of Genova（热那亚大学）；Huawei Technologies Ltd., Ireland Research Center（华为技术有限公司爱尔兰研究中心）；University of Verona（维罗纳大学）
Embracing Collaboration Over Competition: Condensing Multiple Prompts for Visual In-Context Learning
Tsinghua Shenzhen International Graduate School, Tsinghua University（清华大学深圳国际研究生院，清华大学）；Harbin Institute of Technology, Shenzhen（哈尔滨工业大学深圳）；Shenzhen University（深圳大学）；The Hong Kong University of Science and Technology（香港科学与技术大学）；Research Center of Artificial Intelligence, Peng Cheng Laboratory（人工智能研究中心，鹏城实验室）；Meituan, Beijing（美团，北京）
LogicOCR: Do Your Large Multimodal Models Excel at Logical Reasoning on Text-Rich Images?
School of Computer Science, National Engineering Research Center for Multimedia Software, Institute of Artificial Intelligence, and Hubei Key Laboratory of Multimedia and Network Communication Engineering, Wuhan University（计算机学院、多媒体软件国家工程研究中心、人工智能研究院、多媒体与网络通信工程湖北省重点实验室、武汉大学）
