Search

Fine-grained representation is fundamental to species classification based on deep learning, and in this context, cross-modal contrastive learning is an effective method. The diversity of species coupled with the inherent contextual ambiguity of natural language poses a primary challenge in the cross-modal representation alignment of conservation area image data. Integrating cross-modal retrieval tasks with generation tasks contributes to cross-modal representation alignment based on contextual understanding. However, during the contrastive learning process, apart from learning the differences in the data itself, a pair of encoders inevitably learns the differences caused by encoder fluctuations. The latter leads to convergence shortcuts, resulting in poor representation quality and an inaccurate reflection of the similarity relationships between samples in the original dataset within the shared space of features. To achieve fine-grained cross-modal representation alignment, we first propose a residual attention network to enhance consistency during momentum updates in cross-modal encoders. Building upon this, we propose momentum encoding from a multi-task perspective as a bridge for cross-modal information, effectively improving cross-modal mutual information, representation quality, and optimizing the distribution of feature points within the cross-modal shared semantic space. By acquiring momentum encoding queues for cross-modal semantic understanding through multi-tasking, we align ambiguous natural language representations around the invariant image features of factual information, alleviating contextual ambiguity and enhancing model robustness. Experimental validation shows that our proposed multi-task perspective of cross-modal momentum encoders outperforms similar models on standardized image classification tasks and image–text cross-modal retrieval tasks on public datasets by up to 8% on the leaderboard, demonstrating the effectiveness of the proposed method. Qualitative experiments on our self-built conservation area image–text paired dataset show that our proposed method accurately performs cross-modal retrieval and generation tasks among 8142 species, proving its effectiveness on fine-grained cross-modal image–text conservation area image datasets. [ABSTRACT FROM AUTHOR]

Segmented plane mirrors constitute a crucial component in the self-aligned detection process for large-aperture space optical imaging systems. Surface shape errors inherent in segmented plane mirrors primarily manifest as tilt errors and piston errors between sub-mirrors. While the detection and adjustment techniques for tilt errors are well-established, addressing piston errors poses a more formidable challenge. This study introduces a novel approach to achieve long-range, high-precision, and efficient co-phase detection of segmented plane mirrors by proposing a segmented plane mirror shape detection method based on grazing incidence interferometry. This method serves to broaden the detection range of piston errors, mitigate the issue of the 2 π ambiguity resulting from piston errors in co-phase detection, and extend the detection capabilities of the interferometer. By manipulating the incident angle of the interferometer, both rough and precise adjustments of the segmented plane mirrors can be effectively executed. [ABSTRACT FROM AUTHOR]