Meta-DETR: Image-Level Few-Shot Object Detection with Inter-Class Correlation Exploitation (2103.11731v3)

Abstract: Few-shot object detection has been extensively investigated by incorporating meta-learning into region-based detection frameworks. Despite its success, the said paradigm is constrained by several factors, such as (i) low-quality region proposals for novel classes and (ii) negligence of the inter-class correlation among different classes. Such limitations hinder the generalization of base-class knowledge for the detection of novel-class objects. In this work, we design Meta-DETR, a novel few-shot detection framework that incorporates correlational aggregation for meta-learning into DETR detection frameworks. Meta-DETR works entirely at image level without any region proposals, which circumvents the constraint of inaccurate proposals in prevalent few-shot detection frameworks. Besides, Meta-DETR can simultaneously attend to multiple support classes within a single feed-forward. This unique design allows capturing the inter-class correlation among different classes, which significantly reduces the misclassification of similar classes and enhances knowledge generalization to novel classes. Experiments over multiple few-shot object detection benchmarks show that the proposed Meta-DETR outperforms state-of-the-art methods by large margins. The implementation codes will be released at https://github.com/ZhangGongjie/Meta-DETR.

• The paper introduces Meta-DETR, a novel DETR-based framework that leverages inter-class correlation to tackle few-shot object detection challenges.
• It employs meta-learning strategies combined with attention mechanisms to integrate global context and object relationships effectively.
• Extensive experiments demonstrate that Meta-DETR achieves competitive accuracy and robustness compared to state-of-the-art methods with minimal training samples.

Overview of AAAI Press Formatting Instructions for Authors Using LaTeX

The paper, authored by the AAAI Press Staff with contributions from Pater Patel Schneider et al., provides a comprehensive guide for authors using LaTeX to format papers for submission to AAAI Press. This document is critical for ensuring consistency and quality in the presentation of academic papers within AAAI proceedings.

Formatting Requirements

The paper details the essential formatting rules stipulated by AAAI Press, emphasizing uniformity in appearance. Authors are required to use the 2022 AAAI Press LaTeX style files (aaai22.sty, aaai22.bst) explicitly supplied in the AAAI Author Kit. It imposes strict adherence to formatting elements such as font type, size, margins, and paper size, specifically mandating Times Roman or Nimbus fonts and US letter paper size.

Compliance and Submission

An important aspect of the guidelines is the compliance mechanism. Authors must ensure that their .tex files compile without error in PDFLaTeX, adhere to mandatory metadata inclusion, and forbid specific commands that alter layout properties. Files not conforming to these standards will be returned for correction. The paper highlights the necessity of supplying all necessary components such as a fully compliant PDF file, a single .tex source file, and the bibliography file.

Technical Specificities

The document explores the technicalities of formatting. For instance, it prohibits the use of Type 3 fonts, requires the embedding of all fonts in the PDF, and insists on the use of CMYK color space instead of RGB. Additionally, there is a comprehensive list of unsupported packages and commands that, if used, could lead to the rejection of the submission.

Practical Implications

The practical implications of these guidelines are significant for authors preparing submissions. Ensuring compliance not only facilitates a smoother publication process but also promotes the professional presentation required by AAAI standards. Moreover, the document's exhaustive nature serves as a critical resource for troubleshooting and adhering to conference-specific submission requirements.

Future Outlook

As LaTeX remains a staple in academic paper preparation, continuous iterations of such guidelines are anticipated, potentially incorporating adjustments to accommodate evolving typesetting technologies and addressing any emerging challenges authors face. The emphasis on standardization is likely to persist, ensuring high-quality presentation across submissions.

Conclusion

The paper, while technical, is indispensable for authors aiming to publish with AAAI Press. The meticulous detail provided serves to assist authors in navigating the formatting and submission process effectively, thus contributing to the overall excellence and uniformity of AAAI publications.