Joakim Bruslund Haurum

I am a Postdoctoral Researcher in the Visual Analysis and Perception Laboratory at Aalborg University, where I work on computer vision and fine-grained visual categorization. I'm also an affiliated member of the Pioneer Centre for Artificial Intelligence and an ELLIS Member. Previously, I was a member of the Young Academy Panel at the Danish Data Science Academy.

During my Ph.D. I worked on Computer Vision Aided Sewer Inspections and Marine Vision, advised by Professor Thomas B. Moeslund. I have previously visited the Human Pose Recovery and Behavior Analysis in 2021 working with Professor Sergio Escalera, been a visiting researcher at the Vector Institute and the Machine Learning Research Group at the University of Guelph in 2023 working with Professor Graham Taylor, and in 2024 I was a visiting researcher at the University of Edinburgh working together with Reader Oisin Mac Aodha.

Email  /  Google Scholar  /  Github  /  HuggingFace  /  Twitter  /  LinkedIn  /  Headshot

profile photo
What's New
[Sep 2024] Back from my paternity leave. While I was gone our ATC paper was accepted at ECCV and I become an ELLIS member!
[May 2024] Back from my research stay in Edinburgh and now on 3 month parental leave. See you in September!
[Feb 2024] Started as a Visiting Researcher at the University of Edinburgh working with Lecturer Oisin Mac Aodha for the next 3.5 months!
[Sep 2023] Got our BIOSCAN-1M Insect Dataset accepted at NeurIPS 2023!
[Aug 2023] Our in-depth analysis of Token Reduction methods in ViTs is accepted at the "New Ideas in Vision Transformers" ICCV workshop.
[Apr 2023] Gave a talk at York University. Thank you for the invite Kosta!
[Feb 2023] Presented a poster at the Vector Institute Research Symposium 2023.
[Jan 2023] Started as a Research Intern at the Vector Institute working with Prof. Graham Taylor for the next 3 months.
[Nov 2022] Attended DDS and co-hosted the MLOps and Reproducible AI parallel session.
[Oct 2022] Presented MOTCOM at ECCV and hosted the CVCIE workshop.
[Sep 2022] Started as Post Doc. in VAP lab with funding from the Pioneer Centre for AI.
[Jul 2022] Attended ICVSS in sunny Sicily!
[Jun 2022] Attended CVPR and co-hosted the CVSports workshop.
[Jun 2022] Defended my Ph.D.! Many thanks to my committee for a great defence.
[Apr 2022] Submitted my Ph.D. thesis.
[Mar 2022] Attended the opening of the Pioneer Centre for AI and co-hosted the Workshop on Climate and Conservation.
[Jan 2022] Presented our recent work on mutli-task classification at WACV.
Research

I'm interested in computer vision, machine learning, marine vision, fine-grained visual categorization, and real-life applications. Representative papers are highlighted.

An Empirical Study into Clustering of Unseen Datasets with Self-Supervised Encoders
Scott C. Lowe*, Joakim Bruslund Haurum*, Sageev Oore**, Thomas B. Moeslund**, Graham W. Taylor**
Under Review, 2024
arXiv / code / NeurIPS SSL Workshop / NeurIPS R0-FOMO Workshop / ICML FM-Wild Workshop

We explore whether pretrained models can provide a useful representation space for datasets they were not trained on for the purpose of grouping novel unlabelled data into meaningful clusters.

BIOSCAN-CLIP: Bridging Vision and Genomics for Biodiversity Monitoring at Scale
ZeMing Gong, Austin Wang, Joakim Bruslund Haurum, Scott C. Lowe, Graham W. Taylor, Angel X. Chang
Under Review, 2024
arXiv / code / CVPR FGVC Workshop

We introduce BIOSCAN-CLIP, the first approach to use contrastive learning for aligning biological images with DNA barcodes and taxonomic labels to enhance taxonomic classification.

BIOSCAN-5M: A Multimodal Dataset for Insect Biodiversity
Zahra Gharaee*, Scott C. Lowe*, ZeMing Gong*, Pablo Millan Arias*, Nicholas Pellegrino, Austin T. Wang, Joakim Bruslund Haurum, Iuliia Zarubiieva, Lila Kari, Dirk Steinke**, Graham W. Taylor**, Paul Fieguth**, Angel X. Chang**
NeurIPS (D&B Track), 2024
arXiv / code

We construct a large-scale fine-grained dataset of the Insect class with 5M data samples, each containing a biological taxonomic annotation, DNA barcode sequence, geographical location, and RGB image.

Agglomerative Token Clustering
Joakim Bruslund Haurum, Sergio Escalera, Graham W. Taylor, Thomas B. Moeslund
ECCV, 2024
arXiv / code / models / bibtex

We present Agglomerative Token Clustering (ATC), a novel token merging method that consistently outperforms previous token merging and pruning methods across image classification, image synthesis, and object detection & segmentation tasks.

From NeRF to 3DGS: A Leap in Stereo Dataset Quality?
Magnus Kaufmann Gjerde, Filip Slezák, Joakim Bruslund Haurum, Thomas B. Moeslund
CVPR Workshops, 2024
bibtex

We investigate whether using 3D Gaussian Splatting (3DGS) instead of NeRFs to produce stereo camera views leads to better dense disparity labels.

Enhancing Direct Visual Odometry with Deblurring and Saliency Maps
Magnus Kaufmann Gjerde, Kamal Nasrollahi, Thomas B. Moeslund, Joakim Bruslund Haurum
ICMVA, 2024
bibtex

We investigate the effect of integrating a deblurring module with a saliency predictor to perform better point sampling for direct visual odometry.

AssemblyNet: A Point Cloud Dataset and Benchmark for Predicting Part Directions in an Exploded Layout
Jesper Gaarsdal*, Joakim Bruslund Haurum*, Sune Wolff, Claus Brøndgaard Madsen
WACV, 2024
code / bibtex

We propose AssemblyNet, a novel dataset for predicting part directions in assembly models for exploded view visuzalitions. We propose a novel two-path network for predicting part directions.

BarcodeBERT: Transformers for Biodiversity Analysis
Pablo Millan Arias*, Niousha Sadjadi*, Monireh Safari*, ZeMing Gong**, Austin T. Wang**, Scott C. Lowe, Joakim Bruslund Haurum, Iuliia Zarubiieva, Dirk Steinke, Lila Kari, Angel X. Chang, Graham W. Taylor
NeurIPS SSL Workshop, 2023
arXiv / code / bibtex

We propose BarcodeBERT, the first self-supervised method for general biodiversity analysis and highlight how the role of self-supervised pretraining in achieving high-accuracy DNA barcode-based identification at the species and genus level.

A Step Towards Worldwide Biodiversity Assessment: The BIOSCAN-1M Insect Dataset
Zahra Gharaee*, ZeMing Gong*, Nicholas Pellegrino*, Iuliia Zarubiieva, Joakim Bruslund Haurum, Scott C. Lowe, Jaclyn T.A. McKeown, Chris C.Y. Ho, Joschka McLeod, Yi-Yun C Wei, Jireh Agda, Sujeevan Ratnasingham, Dirk Steinke**, Angel X. Chang**, Graham W. Taylor**, Paul Fieguth**
NeurIPS (D&B Track), 2023
project page / arXiv / code / bibtex

We construct a large-scale fine-grained dataset of the Insect class with 1M data samples, each containing a biological taxonomic annotation, DNA barcode sequence, and RGB image.

Which Tokens to Use? Investigating Token Reduction in Vision Transformers
Joakim Bruslund Haurum, Sergio Escalera, Graham W. Taylor, Thomas B. Moeslund
ICCV Workshops, 2023
project page / arXiv / code / bibtex

We conduct the first systematic comparison and analysis of 10 state-of-the-art token reduction methods across four image classification datasets, trained using a single codebase and consistent training protocol.

Multi-scale hybrid vision transformer and Sinkhorn tokenizer for sewer defect classification
Joakim Bruslund Haurum, Meysam Madadi, Sergio Escalera, Thomas B. Moeslund
AiC, 2022
project page / code / bibtex

We propose a novel method for to model spatial semantics in images, where features are aggregated at different scales non-locally using a lightweight vision transformer, and novel Sinkhorn clustering-based tokenizer.

MOTCOM: The Multi-Object Tracking Dataset Complexity Metric
Malte Pedersen, Joakim Bruslund Haurum, Patrick Dendorfer, Thomas B. Moeslund
ECCV, 2022
project page / arXiv / code / bibtex

We propose a set of metrics to determine the complexity of Multi-Object Tracking datasets along three axes: Motion, Appearance, and Occlusion.

A Deep Dive into Computer Vision Aided Sewer Inspections
Joakim Bruslund Haurum
Ph.D. Thesis, 2022
bibtex

My Ph.D. thesis summarizing my work on computer vision aided sewer inspections. Supervised by Thomas B. Moeslund.

Assessment committe: Georgios A. Triantafyllidi (chair.), Serge Belongie, and Graham W. Taylor.

Re-Identification of Giant Sunfish using Keypoint Matching
Malte Pedersen, Joakim Bruslund Haurum, Thomas B. Moeslund, Marianne Nyegaard
NLDL, 2022
bibtex

We propose a keypoint matching based framework for re-identification of Giant Sunfish, comparing SIFT, RootSIFT, and SuperPoint descriptors.

Multi-Task Classification of Sewer Pipe Defects and Properties using a Cross-Task Graph Neural Network Decoder
Joakim Bruslund Haurum, Meysam Madadi, Sergio Escalera, Thomas B. Moeslund
WACV, 2022
project page / arXiv / code / bibtex

We propose a novel decoder-focused multi-task classification architecture called Cross-Task Graph Neural Network (CT-GNN), which refines the disjointed per-task predictions using cross-task information.

Sewer-ML: A Multi-Label Sewer Defect Classification Dataset and Benchmark
Joakim Bruslund Haurum, Thomas B. Moeslund
CVPR, 2021
project page / arXiv / code / dataset / bibtex

We present the world's first open-source sewer defect dataset, with multi-label annotations and 1.3 million images from Danish sewers.

Sewer Defect Classification using Synthetic Point Clouds
Joakim Bruslund Haurum, Moaaz M. J. Allahham, Mathias S. Lynge, Kasper Schøn Henriksen, Ivan A. Nikolov, Thomas B. Moeslund
VISAPP, 2021
code / dataset / bibtex

Investigation into comibning synthetic and real point cloud data of sewer pipes for sewer defect classification using PointNet and DGCNN.

Water Level Estimation in Sewer Pipes using Deep Convolutional Neural Networks
Joakim Bruslund Haurum, Chris H. Bahnsen, Malte Pedersen, Thomas B. Moeslund
Water, 2020
code / bibtex

We estimate the water level in sewer pipes by casting the task as a classification and regression problem.

3D-ZeF: A 3D Zebrafish Tracking Benchmark Dataset
Malte Pedersen*, Joakim Bruslund Haurum*, Stefan Hein Bengtson, Thomas B. Moeslund
CVPR, 2020
project page / arXiv / code / dataset / bibtex

We presenta a dataset for 3D multi-object tracking of zebrafish, captured using an off-the-shelf setup.

Generating Synthetic Point Clouds of Sewer Networks: An Initial Investigation
Kasper Schøn Henriksen, Mathias S. Lynge, Mikkel D. B. Jeppesen, Moaaz M. J. Allahham, Ivan A. Nikolov, Joakim Bruslund Haurum, Thomas B. Moeslund
AVR, 2020
code / bibtex

We propose a system for generating synthetic point clouds of sewer pipes using Structured Domain Randomization for the generation of the sewer systems and an approximated model of a Pico Flexx Time-of-Flight camera.

A Survey on Image-Based Automation of CCTV and SSET Sewer Inspections
Joakim Bruslund Haurum, Thomas B. Moeslund
AiC, 2020
bibtex

We reviewed 113 articles on image-based automated sewer inspection methodologies, finding that there are is a severe lack of 1) publicly available datasets, 2) commonly agreed upon evaluation protocols, and 3) open-sourced code.

Re-Identification of Zebrafish using Metric Learning
Joakim Bruslund Haurum*, Anastasija Karpova*, Malte Pedersen, Stefan Hein Bengtson, Thomas B. Moeslund
WACV Workshops, 2020
code / dataset / bibtex

We show that it is possible to re-identify zebrafish from a sideview, using metric learning and classical feature descriptors.

Detection of Marine Animals in a New Underwater Dataset with Varying Visibility
Malte Pedersen, Joakim Bruslund Haurum, Rikke Gade, Thomas B. Moeslund, Niels Madsen
CVPR Workshops, 2019
dataset / bibtex

We present a new publicly available underwater dataset with annotated image sequences of fish, crabs, and starfish captured in brackish water with varying visibility.

Is it Raining Outside? Detection of Rainfall using General-Purpose Surveillance Cameras
Joakim Bruslund Haurum, Chris H. Bahnsen, Thomas B. Moeslund
CVPR Workshops, 2019
arXiv / code / dataset / bibtex

We design a system for the detection of rainfall by the use of surveillance cameras, and compare it against the former state-of-the-art method for rain detection on our new AAU Visual Rain Dataset (VIRADA).

Pixel Reprojection of 360 Degree Renderings for Small Parallax Effects
Joakim Bruslund Haurum, Christian Nygaard Daugbjerg, Péter Rohoska, Andrea Coifman, Anne Juhler Hansen, Martin Kraus
AVR, 2017
bibtex

We apply pixel reprojection on nine 360 degree renderings to enable 3D motion and introduce motion parallax effects, without explicit knowledge of the 3D geometry.

AAUSAT5 - an evaluation of a student-run cubesat project
Rasmus Gundorff Sæderup, Joakim Bruslund Haurum
SSEA, 2015
bibtex

We evaluate and discuss the experiences acquired during the 18 month long production, qualification and testing phase of the AAUSAT5 cubesat.


This website layout was adapted from Jon Barron's template.