MétaCan
Menu
Back to cohort
Record W7071587911

Self-Supervised Learning for Semantic Segmentation of Images

2023· dissertation· en· W7071587911 on OpenAlexfundno aff

Bibliographic record

VenueUniversity Library (University of Saskatchewan) · 2023
Typedissertation
Languageen
FieldComputer Science
TopicDomain Adaptation and Few-Shot Learning
Canadian institutionsnot available
FundersCanada First Research Excellence Fund
KeywordsExploitSegmentationPascal (unit)Redundancy (engineering)Artificial neural networkImage segmentationDeep learningTask (project management)Multi-task learning
DOInot available

Abstract

fetched live from OpenAlex

Artificial Neural Networks (ANN) are powerful Machine Learning (ML) models that can help solve problems that are hard or even impossible to design solutions for by hand. These models learn to exploit\ninformation present in their target datasets to solve various problems. However, labelling data can be quite\nexpensive, time-consuming and often requires a domain expert. Therefore it would be quite beneficial if one\ncould train a model in such a way that exploits unlabeled data. Fortunately, Self-Supervised Learning (SSL)\nmethods are a family of learning algorithms that attempt to do just that. Many SSL methods exist, but in\nthis thesis, we explore Barlow Twins (BT) — a siamese network based on redundancy reduction, and Image\nReconstruction (IR) — a method proposed in Karnam’s thesis. In addition, we extend the Image Reconstruction method with both Coarse Cutout and Hide-and-Seek augmentations as they have been applied in similar\nsupervised and weakly-supervised segmentation task scenarios. We apply these methods and investigate the\nresults with the PASCAL VOC dataset.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

How this classification was reachedexpand

Full frame distilled prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.

metaresearch head score (Codex)0.000
metaresearch head score (Gemma)0.000
Version: codex-gemma-dda1882f352aValidation status: machine_predicted_unvalidated
Candidate categoriesMeta-epidemiology (narrow)
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Qualitative · Consensus signal: Qualitative
GenreCandidate signal: Empirical · Consensus signal: Empirical
Teacher disagreement score0.195
Threshold uncertainty score1.000

Codex and Gemma teacher scores by category

CategoryCodexGemma
Metaresearch0.0000.000
Meta-epidemiology (narrow)0.0000.000
Meta-epidemiology (broad)0.0000.000
Bibliometrics0.0010.001
Science and technology studies0.0000.000
Scholarly communication0.0000.002
Open science0.0010.000
Research integrity0.0000.000
Insufficient payload (model declined to judge)0.0000.000

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.009
GPT teacher head0.192
Teacher spread0.183 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; a candidate call from one teacher head, not a consensus.

Study designQualitative
Domainnot available
GenreEmpirical

How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".

Quick stats

Citations0
Published2023
Admission routes1
Has abstractyes

Explore more

Same venueUniversity Library (University of Saskatchewan)Same topicDomain Adaptation and Few-Shot LearningFrench-language works237,207