Cross-domain Multi-modal Few-shot Object Detection via Rich Text

2024-03-24

arXiv.org

AI Summary

To address severe domain shift and scarce annotated samples—leading to substantial performance degradation in cross-domain few-shot multimodal object detection—this paper proposes a meta-learning-based vision-language collaborative modeling framework. The method introduces, for the first time, a bidirectional text feature generation mechanism to construct a textual semantic calibration module; it achieves fine-grained vision-language feature alignment via joint encoding with BERT and CLIP, thereby facilitating effective cross-domain knowledge transfer. Additionally, the approach integrates multimodal feature aggregation with support-set adaptive modeling. Evaluated on standard cross-domain few-shot detection benchmarks, the method achieves an average precision (mAP) improvement of 12.6% over prior work, demonstrating significantly enhanced model robustness and generalization capability.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Web Mining and Content Analysis: Mining multimedia, multimodal, multilingual, cross-lingual Web data

Abstract

Cross-modal feature extraction and integration have led to steady performance improvements in few-shot learning tasks due to generating richer features. However, existing multi-modal object detection (MM-OD) methods degrade when facing significant domain-shift and are sample insufficient. We hypothesize that rich text information could more effectively help the model to build a knowledge relationship between the vision instance and its language description and can help mitigate domain shift. Specifically, we study the Cross-Domain few-shot generalization of MM-OD (CDMM-FSOD) and propose a meta-learning based multi-modal few-shot object detection method that utilizes rich text semantic information as an auxiliary modality to achieve domain adaptation in the context of FSOD. Our proposed network contains (i) a multi-modal feature aggregation module that aligns the vision and language support feature embeddings and (ii) a rich text semantic rectify module that utilizes bidirectional text feature generation to reinforce multi-modal feature alignment and thus to enhance the model's language understanding capability. We evaluate our model on common standard cross-domain object detection datasets and demonstrate that our approach considerably outperforms existing FSOD methods.

Problem

• Cross-domain object detection
• Few-shot learning
• Multi-modal feature alignment

Innovation

• Utilizes rich text semantic information
• Multi-modal feature aggregation module
• Bidirectional text feature generation