About

I am a Senior Staff Research Scientist / Tech Lead Manager @ Google DeepMind, working on multimodal AI (computer vision, machine learning and related areas). I lead a team of researchers / engineers split between Mountain View (US) and Sao Paulo (Brazil).

I graduated with a Ph.D. from Stanford University in 2016. I am a Senior Member of the IEEE, and was previously also a Fulbright Science & Technology Scholar and an Accel Leadership Scholar.

News

• [Jun’25] I’m giving a talk at the CVPR’25 BEAM workshop, entitled “What’s missing in multimodal AI? Towards spatial awareness, effective tool use and fine-grained understanding”.
• [Mar’25] New paper up on arXiv: L. Xie, J. Guo, O. Cakmakci, A. Araujo, L. A. Jeni, Z. Jia, “AlignDiff: Learning Physically-Grounded Camera Alignment via Diffusion”
• [Mar’25] I will serve as Area Chair for NeurIPS’25.
• [Feb’25] CVPR’25 highlight paper: T. Novello*, D. Aldana*, A. Araujo, L. Velho, “Tuning the Frequencies: Robust Training for Sinusoidal Neural Networks”.
• [Feb’25] Our paper on rank losses for image retrieval is now accepted into TPAMI!
• [Jan’25] Our TIPS paper is now accepted into ICLR’25!
• [Dec’24] I will serve as Area Chair for ICCV’25 (Lead AC role).
• [Oct’24] New paper up on arXiv: K. Maninis*, K. Chen*, S. Ghosh*, A. Karpur*, K. Chen, Y. Xia, B. Cao, D. Salz, G. Han, J. Dlabal, D. Gnanapragasam, M. Seyedhosseini, H. Zhou, A. Araujo, “TIPS: Text-Image Pretraining with Spatial awareness”
• [Sep’24] Our UDON paper is now accepted into NeurIPS’24!
• [Sep’24] I will serve as Area Chair for CVPR’25.
• [Jun’24] New paper up on arXiv: N. Ypsilantis, K. Chen, A. Araujo, O. Chum, “UDON: Universal Dynamic Online distillatioN for generic image representations”
• [Jun’24] The Instance-Level Recognition workshop is back at ECCV’24! And for the first time in 6 editions, with a call for papers. We look forward to your submissions!
• [May’24] Our two CVPR’24 papers are now on arXiv:
  • H. Jiang, A. Karpur, B. Cao, Q. Huang, A. Araujo, “OmniGlue: Generalizable Feature Matching with Foundation Model Guidance”
  • G. Potje, F. Cadar, A. Araujo, R. Martins, E. Nascimento, “XFeat: Accelerated Features for Lightweight Image Matching”
• [May’24] I will serve as Area Chair for NeurIPS’24.
• [Apr’24] New paper up on arXiv: L. Castrejon, T. Mensink, H. Zhou, V. Ferrari, A. Araujo, J. Uijlings, “HAMMR: HierArchical MultiModal React agents for generic VQA”
• [Feb’24] 2 papers accepted at CVPR’24! OmniGlue and Xfeat.
• [Nov’23] I will serve as Area Chair for ECCV’24.
• [Oct’23] New papers accepted: NAVI in NeurIPS’23 and LFM-3D as an oral in 3DV’24!
• [Oct’23] New post on the Google Brazil blog about recent AI research from Google (in Portuguese).
• [Sep’23] 4 papers accepted at ICCV’23! Encyclopedic-VQA, CANN, SuperGlobal and Universal Embeddings.
• [Sep’23] Three new papers up on arXiv:
  • S. Shao*, K. Chen, A. Karpur, Q. Cui, A. Araujo, B. Cao*, “Global Features are All You Need for Image Retrieval and Reranking” (ICCV’23)
  • N. Ypsilantis, K. Chen, B. Cao, M. Lipovský, P. Dogan-Schönberger, G. Makosa, B. Bluntschli, M. Seyedhosseini, O. Chum, A. Araujo, “Towards Universal Image Embeddings: A Large-Scale Dataset and Challenge for Generic Image Representations” (ICCV’23)
  • E. Ramzi, N. Audebert, C. Rambour, A. Araujo, X. Bitot, N. Thome, “Optimization of Rank Losses for Image Retrieval”
• [Jul’23] I will serve as Area Chair for CVPR’24.
• [Jun’23] Three new papers up on arXiv:
  • D. Aiger, A. Araujo, S. Lynen, “Yes, we CANN: Constrained Approximate Nearest Neighbors for local feature-based visual localization” (ICCV’23)
  • T. Mensink*, J. Uijlings*, L. Castrejon, A. Goel, F. Cadar, H. Zhou, F. Sha, A. Araujo, V. Ferrari, “Encyclopedic VQA: Visual questions about detailed properties of fine-grained categories” (ICCV’23)
  • V. Jampani*, K. Maninis*, A. Engelhardt, A. Karpur, K. Truong, K. Sargent, S. Popov, A. Araujo, R. Martin-Brualla, K. Patel, D. Vlasic, V. Ferrari, A. Makadia, C. Liu, Y. Li, H. Zhou, “NAVI: Category-Agnostic Image Collections with High-Quality 3D Shape and Pose Annotations”
• [Mar’23] New paper up on arXiv: A. Karpur, G. Perrotta, R. Martin-Brualla, H. Zhou, A. Araujo, “LFM-3D: Learnable Feature Matching Across Wide Baselines Using 3D Signals”
• [Feb’23] CVPR’23 paper: G. Potje, F. Cadar, A. Araujo, R. Martins, E. Nascimento, “Enhancing Deformable Local Features by Jointly Learning to Detect and Describe Keypoints”. See also project page
• [Sep’22] I will serve as Area Chair for CVPR’23.
• [Jul’22] Universal Embedding Challenge launched! We welcome everyone’s participation! See also the Google AI blog post.
• [Mar’22] Instance-Level Recognition 2022 workshop accepted at ECCV’22, with new upcoming exciting challenges on universal image embeddings and language-assisted product retrieval. Feb’23 update: all workshop recordings are on the website!
• [Jan’22] Our DELF/DELG/Receptive-field papers and instance-level recognition workshop were cited in Prof. Szeliski’s new computer vision textbook version.
• [Oct’21] BMVC’21 paper: A. Iscen, A. Araujo, B. Gong and C. Schmid, “Class-Balanced Distillation for Long-Tailed Visual Recognition”.
• [Aug’21] New paper up on arXiv: Z. Kim, A. Araujo, B. Cao, C. Askew, J. Sim, M. Green, N. F. Tilla and T. Weyand, “Towards A Fairer Landmark Recognition Dataset”.
• [May’21] Instance-Level Recognition 2021 workshop accepted at ICCV’21, see landmark retrieval and recognition challenges.
• [Oct’20] I gave a keynote talk at the SUMAC workshop at ACM MM 2020, see the video recording here.
• [Sep’20] Check out our Google AI blog post featuring DELG, our open-source codebase, landmark recognition challenges and the ECCV Instance-level Recognition workshop.
• [Jul’20] ECCV’20 paper: B. Cao*, A. Araujo* and J. Sim, “Unifying Deep Local and Global Features for Image Search”.
• [Apr’20] CVPR’20 paper (oral): T. Weyand*, A. Araujo*, B. Cao and J. Sim, “Google Landmarks Dataset v2 - A Large-Scale Benchmark for Instance-Level Recognition and Retrieval”.
• [Feb’20] Instance-Level Recognition workshop accepted at ECCV’20, see landmark retrieval and recognition challenges.
• [Nov’19] Distill paper: A. Araujo, W. Norris and J. Sim, “Computing Receptive Fields of Convolutional Neural Networks”
• [Jun’19] DELF achieved 2nd place in CVPR Visual Localization challenge (Local Features track). See our slides here.
• [Jun’19] CVPR’19 paper: M. Teichmann*, A. Araujo*, M. Zhu and J. Sim, “Detect-to-Retrieve: Efficient Regional Aggregation for Image Search”. Poster here, code here.
• [Dec’18] 2nd Landmark Recognition workshop accepted at CVPR’19. Hope you can participate in the challenges: Landmark Recognition and Landmark Retrieval.
• [Mar’19] MIPR’19 paper (Best Paper Award!): J.-B. Boin, A. Araujo and B. Girod, “Recurrent Neural Networks for Person Re-identification Revisited”
• [Mar’18] Google Research blog post discussing the Google-Landmarks dataset and challenge
• [Dec’17] We are organizing the Landmark Recognition Challenge and the Landmark Retrieval Challenge, associated to the Landmark Recognition CVPR’18 workshop. Hope you can participate!
• [Oct’17] See the DELF project webpage for code and dataset.