1st place out of 588
in
Youth Mental Health Narratives: Automated Abstraction
(18,000US$)
Developed a transformer model to automate data abstraction from youth suicide
narratives
code
natural language processing, classification, health
Published on 2024-12-25
1st place out of 50
in Machine Learning track of
Yandex Cup
(500,000RUB≈5,500US$)
Built a gesture-typing prediction model using bidirectional LSTM model
code |
video presentation |
in press |
archive en |
archive ru
natural language processing, seq2seq
Published on 2023-11-15
2nd place out of 89
in Machine Learning track of
Yandex Cup
(out of competition due to first place in NeuroSwipe track)
Developed a nowcasting model leveraging time series analysis and machine
learning for short-term forecasting
code |
archive en |
archive ru
computer vision, seq2seq
Published on 2023-11-15
2nd place out of 335
in
AISG–SLA Visual Localisation Challenge
(10,000US$)
Created model for accurate camera pose estimation from 2D monocular images,
using computer vision
code
computer vision, regression, monocular visual odometry
Published on 2023-8-16
1st place out of 976
in
The BioMassters
(5,000US$)
Estimated yearly forest biomass across Finland using satellite imagery
(Sentinel-1/2) and deep learning techniques
code
computer vision, aerial images, geospatial images, regression, height estimation
Published on 2023-1-28
2nd place out of 321
in
Deep Chimpact: Depth Estimation for Wildlife Conservation
(2,000US$)
Developed a deep learning model to estimate the distance to animals in camera
trap videos
code
computer vision, video, regression
Published on 2021-11-16
2nd place out of 47
in
Machine Learning challenge Yandex Cup (Computer Vision)
Develop a zero-shot model for image classification
code
computer vision, images, classification, zero-shot learning
Published on 2021-10-20
2nd place out of 54
in
Machine Learning challenge Yandex Cup (Automatic Speech Recognition)
Develop a speech recognition model for noisy data
code
speech recognition, audio, classification
Published on 2021-10-20
4th place out of 768
in
Overhead Geopose Challenge
(3,000US$)
Improved the utility of satellite imagery for time-critical applications
(disaster response) through deep learning
code
computer vision, aerial images, geospatial images, regression, height estimation
Published on 2021-07-21
3rd place out of 29
in
Waymo Open Dataset Challenge 1: Motion Prediction
(666US$)
Predicted the future positions of agents based on historical trajectory data,
utilizing convolutional neural networks
code |
paper on arXiv |
mass media en |
mass media ru
computer vision, images, regression, self-driving cars
Published on 2021-06-20
3rd place out of 547
in
TissueNet: Detect Lesions in Cervical Biopsies
(8,000€≈9400US$)
Built a convolutional neural network to detect severe epithelial lesions in
cervical biopsy images, assisting in medical diagnosis
code |
talk |
presentation |
mass media
computer vision, high-res medical images, classification
Published on 2020-11-15
2nd place out of 922
in
Clog Loss: Advance Alzheimer’s Research with Stall Catchers
(2,000US$)
Developed a machine learning model to classify blood vessel flow in brain
imagery, aiding research into Alzheimer's disease
code |
bot author
computer vision, medical videos, classification
Published on 2020-08-03
2nd place out of 1106
in
Open Cities AI Challenge: Segmenting Buildings for Disaster Resilience
(4,000US$)
Created a semantic segmentation model to identify building footprints from
aerial imagery, supporting disaster preparedness and response
code |
report
computer vision, aerial images, geospatial images, segmentation
Published on 2020-03-18
1st place out of 140
in
Machine Learning track of Yandex programming championship
(300,000RUB≈4700US$)
Developed a topic model classifier
code |
archive en |
archive ru
natural language processing, text, classification
Published on 2019-10-30
1st place out of 104
in
Machine Learning track of Yandex programming championship
(300,000RUB≈4700US$)
Built a recommender system
code |
talk |
archive en |
archive ru
recommender system
Published on 2019-07-10