Benchmarking Zero-Shot Recognition with Vision-Language Models: Challenges on Granularity and Specificity

Documenti analoghi

• Self-Supervised Multi-Object Tracking with Path Consistency
  di: Lu, Zijia, et al.
  Pubblicazione: (2024)
• Beyond 3D VQAs: Injecting 3D Spatial Priors into Vision-Language Models for Enhanced Geometric Reasoning
  di: Yeh, Chun-Hsiao, et al.
  Pubblicazione: (2026)
• Hyperbolic Learning with Synthetic Captions for Open-World Detection
  di: Kong, Fanjie, et al.
  Pubblicazione: (2024)
• Invert2Restore: Zero-Shot Degradation-Blind Image Restoration
  di: Chihaoui, Hamadi, et al.
  Pubblicazione: (2025)
• Zero-Shot Chinese Character Recognition with Hierarchical Multi-Granularity Image-Text Aligning
  di: Zhu, Yinglian, et al.
  Pubblicazione: (2025)