2021/06/24 by Ke-Han Lu, Bo-Han Fang, Lu, Ke-Han +4 · 1 citation
Computer Science · Engineering · #Advanced Image and Video Retrieval Techniques #Adversarial system #Artificial intelligence #Computation and Language (cs.CL) #Computer Vision and Pattern Recognition (cs.CV) #Computer science #Domain Adaptation and Few-Shot Learning #Engineering #FOS: Computer and information sciences #Machine learning #Modal #Multimodal Machine Learning Applications #Set (abstract data type) #Test set #Training set #Transformer #Voltage #cs.CL #cs.CV
paper · pdf · doi:10.48550/arxiv.2106.13033
published in arXiv (Cornell University) (Cornell University) · CVPR 2021 Workshop: Visual Question Answering (VQA) Challenge
openalex publication_date 2021/06/24 · openalex created_date 2021/07/05 · arxiv created 2021/10/26 · arxiv updated 2021/10/27 · openalex updated_date 2026/08/01
In this paper, inspired by the successes of visionlanguage pre-trained models and the benefits from training with adversarial attacks, we present a novel transformerbased cross-modal fusion modeling by incorporating the both notions for VQA challenge 2021. Specifically, the proposed model is on top of the architecture of VinVL model [19], and the adversarial training strategy [4] is applied to make the model robust and generalized. Moreover, two implementation tricks are also used in our system to obtain better results. The experiments demonstrate that the novel framework can achieve 76.72% on VQAv2 test-std set.