Oct 27, 2023 · In this paper, we introduce a new and challenging multi-domain lifelong VQA task, dubbed MDL-VQA, which encourages the VQA model to continuously learn across ...

Nov 3, 2023 · In this paper, we introduce a new and challenging multi-domain lifelong VQA task, dubbed MDL-VQA, which encourages the VQA model to continuously ...

SCD [24] proposes to transfer domain knowledge through self-critical distillation at both the logits and feature levels. It defines instance-relevant and domain ...

ACM International Conference on Multimedia (MM), 2023. Multi-Domain Lifelong Visual Question Answering via Self-Critical Distillation Mingrui Lao, Nan Pu ...

Jun 27, 2024 · Multi-domain lifelong visual question answering via self-critical distillation. In Proceedings of the 31st ACM International Conference on ...

Multi-Domain Lifelong Visual Question Answering via Self-Critical Distillation ... VQA-BC: Robust Visual Question Answering Via Bidirectional Chaining.

Multi-Domain Lifelong Visual Question Answering via Self-Critical Distillation · Author Picture Mingrui Lao. Leiden University, Leiden, Netherlands · Author ...

Multi-Domain Lifelong Visual Question Answering via Self-Critical Distillation ... Multi-Stage Hybrid Embedding Fusion Network for Visual Question Answering

Nov 11, 2019 · This paper proposes a novel supervised multi-modal domain adaptation method for VQA to learn joint feature embeddings across different ...

COCA: COllaborative CAusal Regularization for Audio-Visual Question Answering. ... Multi-Domain Lifelong Visual Question Answering via Self-Critical Distillation.