DRAN: Detailed Region-Adaptive Normalization for Conditional Image Synthesis

Lyu, Yueming; Pei-bin, Chen; Sun, Jian; Peng, Bo; Wang, Xu

doi:10.1109/tmm.2023.3290481

Cited by 3 publications

(1 citation statement)

References 47 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Trained on a vast dataset of 400 million (image, text) pairs, it strives to learn a joint semantical space for both images and texts using contrastive loss. Benefiting from its excellent image/text representation ability, CLIP has found extensive applications in different areas, such as domain adaptation [32], image segmentation [33], [34], image generation and editing [12], [13], [35], [36], [37], [38], [39], [40]. In particular, LAFITE [41] and KNN-Diffusion [42] utilize the CLIP image-text feature space and exploit languagefree text-to-image generation.…”

Section: Vision-language Representationsmentioning

confidence: 99%

A Similarity Distillation Guided Feature Refinement Network for Few-Shot Semantic Segmentation

Lyu

Liu

Chen

et al. 2022

2022 IEEE International Conference on Image Processing (ICIP)

View full text Add to dashboard Cite

Few-shot semantic segmentation is a challenging task of predicting object categories in pixel-wise with only few annotated samples. Existing methods mainly have two problems, which are representation inconsistency of query and support images and semantic-level feature insufficient. To tackle the two problems, we propose a similarity distillation guided feature refinement network (SD-FRNet). Specifically, we first use support label to generate support similarity feature map and coarse prediction of query image. Then, we use this coarse prediction to generate query similarity feature. To compensate feature representation inconsistency, we conduct knowledge distillation mechanism to align similarity features of query and support images. To enrich semantic-level feature, we further design a feature refinement module, which achieves high-quality segmentation. Extensive experiments show the effectiveness of SD-FRNet. On benchmark datasets, PASCAL-5 i and COCO-20 i , our proposed SD-FRNet outperform the previous SOTA (state-of-the-art) results.

show abstract