María Alejandra Bravo Sarmiento

María Alejandra Bravo Sarmiento

Postdoc at TUM and Helmholtz Munich

Researcher in Artificial Intelligence

I am a Post-Doctoral researcher in the Explainable Machine Learning group at TUM and Helmholtz Munich, led by Prof. Zeynep Akata.

I received my PhD in the Computer Vision group at the University of Freiburg, Germany, under the supervision of Prof. Thomas Brox. Previous website.

Research interests

My research focuses on advancing vision-language models for open-vocabulary recognition and generative evaluation. I am particularly interested in the intersection of computer vision and natural language processing, with a focus on developing models that can understand and generate visual content in a more human-like manner.

  • Vision-Language Understanding: Integration of visual and linguistic information for multimodal reasoning.
  • Open-Vocabulary Recognition: Recognition of objects and attributes beyond predefined categories, including fine-grained attributes like color, texture, and material.
  • Human-Aligned AI: Development of AI systems that align with human preferences for improved interaction and collaboration, including building user studies to validate this alignment.
  • AI Generation: Advancing image and text generation, with a focus on aligning both modalities effectively.
  • Dataset, benchmark, and evaluation: Creating datasets and benchmarks for evaluating the performance of vision-language models.
  • Weakly-Supervised and Multimodal Learning: Leveraging weakly supervised learning for grounding visual and textual data

Quick Facts

  • 🎓 PhD in Computer Science, focused on visual-language models
  • 📍 Currently based in Munich, Germany
  • 🧑‍🏫 Passionate about teaching and mentoring students
  • 🎙️ Presented at conferences: ICLR, CVPR and NeurIPS
  • 📝 Reviewed at NeurIPS, CVPR, ICCV, and TPAMI
  • 🤝 Collaborated with researchers from Amazon, KAUST, and the University of Freiburg
  • 🇨🇴 Colombian, with a background in mathematics and biomedical engineering
  • 🎾💃 Enjoys playing tennis and dancing lindy hop and salsa
  • 🗣️ I speak Spanish, English, German and a basic French

Selected Publications

See my Google Scholar profile for a complete list of my publications. Check out my thesis, which provides a detailed compilation of my Ph.D. work:

  • Thesis: Advancing vision-language models for open-vocabulary recognition and generative evaluation (May 2025)

  • TIAlign: Text-Image Concept Human Alignment (WiML NeurIPS Workshop 2024 - Best poster award TWiML 2024)
  • oVQA: Open-ended Visual Question Answering (Spotlight ICLR 2024)
  • OVAD: Open-vocabulary Attribute Detection Dataset and Benchmark (CVPR 2023)
  • LocOV: Localized Vision-Language Matching for Open-vocabulary Object Detection (GCPR 2022)

Biography

I am a Colombian researcher with a background in computer vision, artificial intelligence and biomedical engineering.

I hold a PhD in Computer Science from the University of Freiburg, Germany, where I am part of the Computer Vision Group led by Prof. Thomas Brox. My research focuses on vision-language models, particularly in open-vocabulary recognition and generative evaluation.

During my PhD I did an internship at Amazon in Tübingen, where I worked on Vision-Language alignment and Generative AI with Betty Mohler and Ali Jahanian.

During 2019 I had the opportunity to work as a research intern at KAUST at the Image and Video Understanding Laboratory (IVUL), working on Video Object Segmentation.

Before that I obained my Master’s degree at the Biomedical Computer Vision Group led by Pablo Arbeláez at the Universidad de los Andes in Bogotá, Colombia.

I hold a Bachelor’s degree in Mathematics and another in Biomedical Engineering from the Universidad de los Andes in Bogotá, Colombia.

Scholarships and Grants

  • Best Poster Award at the TWiML 2024 workshop presented at NeurIPS WiML. Outstanding Reviewer Award at NeurIPS D&B in 2023 and 2022.
  • DAAD Research Grants for Doctoral Programs in Germany (2019/20), grant number 57440921. German-Colombian Academic Cooperation between the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation, grant BR 3815/9-1) and the Universidad de los Andes in Colombia.
  • Beca YERLY. Mathematical Program Scholarship for Academic Excellence (2012). Scholarship for Academic Excellence (2012).

Teaching

Student Supervision

I have supervised several students in their Bachelor thesis at the University of Freiburg

  • Improving Visual Grouping and Visual-Text Alignment for Open-Vocabulary Segmentation (Ayushi Sharma, 2023) - Supervision together with Silvio Galesso
  • Multimodal attribute learning (Anna Stroganova, 2022)
  • Improving Clip-Sentence Retrieval with COOT using large-scale noisy-aligned Training Data (Felix Jablonksi, 2022)
  • Applying Hierarchical Representations from Video Retrieval to Video Captioning (Simon Ging, 2021)

Teaching Assistant

  • Co-Organizer of the Deep Learning Lab (2021 and 2022 Freiburg)
  • Assistant and Supervisor of the Deep Learning and Computer Vision Seminars (2019 - 2023 Freiburg)
  • Assistant of Image Analysis and Processing (2018 Universidad de los Andes, Colombia)
  • Lecturer of Linear Algebra and Integral Calculus and Differential Equations (before 2018 Universidad de los Andes, Colombia)