Fernando Julio Cendra

News

• [ Apr. 2026 ] One paper about category discovery with part-level correspondence priors has been accepted to ICML 2026.
• [ Mar. 2026 ] Obtained my MPhil degree at The University of Hong Kong.
• [ Jan. 2026 ] Started my research internship at NVIDIA.
• [ Jun. 2025 ] Our work ICE, CVPR 2025 was featured by HKU Graduate School News Letter, Awards & Achievements.
• [ Feb. 2025 ] One paper about concept learning with diffusion models has been accepted to CVPR 2025 as Highlight.
• [ Jul. 2024 ] One paper about continual category discovery has been accepted to ECCV 2024.
• [ Oct. 2023 ] Honored to receive the Excellent Research Award from my department.
• [ Jun. 2023 ] Our CVPR 2023 work was featured by RSIP Vision Magazine.
• [ Feb. 2023 ] One paper about heterogeneous graph representation learning has been accepted to CVPR 2023.
• [ Dec. 2022 ] One paper about 3D representation learning has been accepted to NeurIPS Workshop 2022.

Publications

* denotes equal contribution

Open-World Visual Concept Discovery and Extraction

Fernando Julio Cendra

MPhil Thesis, The University of Hong Kong, 2026

Thesis (coming soon)

Effective Prompt Pool Learning for Continual Category Discovery

Fernando Julio Cendra, Xinghui Li and Kai Han

arXiv preprint 2026

project

paper

arXiv

PartCo: Part-Level Correspondence Priors Enhance Category Discovery

Fernando Julio Cendra and Kai Han

ICML 2026

project

paper

arXiv

code

ICE: Intrinsic Concept Extraction from a Single Image via Diffusion Models

Fernando Julio Cendra and Kai Han

CVPR 2025 Highlight, 2.8% of submissions

project

paper

arXiv

code

poster

PromptCCD: Learning Gaussian Mixture Prompt Pool for Continual Category Discovery

Fernando Julio Cendra, Bingchen Zhao and Kai Han

ECCV 2024

project

paper

arXiv

code

poster

Histopathology Whole Slide Image Analysis with Heterogeneous Graph Representation Learning

Chan Tsai Hor^*, Fernando Julio Cendra^*, Lan Ma, Guosheng Yin and Lequan Yu

CVPR 2023

project

paper

arXiv

code

poster

SL3D: Self-supervised-Self-labeled 3D Recognition

Fernando Julio Cendra, Lan Ma, Jiajun Shen and Xiaojuan Qi

NeurIPS Workshop on "Self-supervised Learning: Theory & Practice" 2022

paper

arXiv

code

poster