PulseExploreJournal ClubDebatesTrendingResearchersJournals
Instagram
HomeExploreJournal ClubTrending
Synapse
⌘+K
Synapse
January 26, 20260 citations

Topology-Guided Semantic Face Center Estimation for Rotation-Invariant Face Detection.

View Full Paper
HKHathai KaewkornChongqing University of Posts and TelecommunicationsLZLifang ZhouChongqing University of Posts and TelecommunicationsWLWeisheng LiChongqing University of Posts and Telecommunications

Key Points

  • The research aims to enhance face detection accuracy under rotational variations, particularly for out-of-plane rotations.
  • Proposed a topology-guided semantic face center estimation method using graph-based landmark relationships.
  • Constructed a rotation-aware face dataset with precise center annotations and balanced rotation angles.
  • Introduced a Hybrid-ViT model that merges CNN features with transformer global context.
  • Developed a hybrid metric that integrates topological geometry and semantic perception for evaluation.
  • Experimental results show superior performance compared to state-of-the-art models in cross-dataset evaluations.
  • Achieved improved accuracy in face center localization under extreme pose variations.

Abstract

Face detection accuracy significantly decreases under rotational variations, including in-plane (RIP) and out-of-plane (ROP) rotations. ROP is particularly problematic due to its impact on landmark distortion, which leads to inaccurate face center localization. Meanwhile, many existing rotation-invariant models are primarily designed to handle RIP, they often fail under ROP because they lack the ability to capture semantic and topological relationships. Moreover, existing datasets frequently suffer from unreliable landmark annotations caused by imperfect ground truth labeling, the absence of precise center annotations, and imbalanced data across different rotation angles. To address these challenges, we propose a topology-guided semantic face center estimation method that leverages graph-based landmark relationships to preserve structural integrity under both RIP and ROP. Additionally, we construct a rotation-aware face dataset with accurate face center annotations and balanced rotational diversity to support training under extreme pose conditions. Next, we introduce a Hybrid-ViT model that fuses CNN spatial features with transformer-based global context and employ a center-guided module for robust landmark localization under extreme rotations. In order to evaluate center quality, we further design a hybrid metric that combines topological geometry with semantic perception for a more comprehensive evaluation of face center accuracy. Finally, experimental results demonstrate that our method outperforms state-of-the-art models in cross-dataset evaluations. Code: https: //github. com/Catster111/TCERIFD.

Ask AI
Helpful
Bookmark
Share
View Full Paper

Cite This Study

Kaewkorn et al. (2026) studied this question.

synapsesocial.com/papers/6977032e722626c4468e82dfhttps://doi.org/10.1109/tip.2026.3654422
Ask AI
Helpful
Bookmark
Share
View Full Paper