MétaCan
Menu
Back to cohort
Record W2976931991 · doi:10.48550/arxiv.1805.11123

Global Sum Pooling: A Generalization Trick for Object Counting with\n Small Datasets of Large Images

2018· preprint· en· W2976931991 on OpenAlex

Why this work is in the frame

A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.

affAt least one author lists a Canadian institution in the pinned OpenAlex snapshot.

Bibliographic record

VenuearXiv (Cornell University) · 2018
Typepreprint
Languageen
FieldComputer Science
TopicVideo Surveillance and Tracking Methods
Canadian institutionsUniversity of Saskatchewan
Fundersnot available
KeywordsOverfittingPoolingComputer scienceGeneralizationArtificial intelligenceConvolutional neural networkInferencePattern recognition (psychology)Focus (optics)Object (grammar)Image (mathematics)Property (philosophy)Machine learningArtificial neural networkMathematics

Abstract

fetched live from OpenAlex

In this paper, we explore the problem of training one-look regression models\nfor counting objects in datasets comprising a small number of high-resolution,\nvariable-shaped images. We illustrate that conventional global average pooling\n(GAP) based models are unreliable due to the patchwise cancellation of true\noverestimates and underestimates for patchwise inference. To overcome this\nlimitation and reduce overfitting caused by the training on full-resolution\nimages, we propose to employ global sum pooling (GSP) instead of GAP or fully\nconnected (FC) layers at the backend of a convolutional network. Although\ncomputationally equivalent to GAP, we show through comprehensive\nexperimentation that GSP allows convolutional networks to learn the counting\ntask as a simple linear mapping problem generalized over the input shape and\nthe number of objects present. This generalization capability allows GSP to\navoid both patchwise cancellation and overfitting by training on small patches\nand inference on full-resolution images as a whole. We evaluate our approach on\nfour different aerial image datasets - two car counting datasets (CARPK and\nCOWC), one crowd counting dataset (ShanghaiTech; parts A and B) and one new\nchallenging dataset for wheat spike counting. Our GSP models improve upon the\nstate-of-the-art approaches on all four datasets with a simple architecture.\nAlso, GSP architectures trained with smaller-sized image patches exhibit better\nlocalization property due to their focus on learning from smaller regions while\ntraining.\n

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.

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)0.001
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: Simulation or modeling · Consensus signal: Simulation or modeling
GenreCandidate signal: Empirical · Consensus signal: none
Teacher disagreement score0.790
Threshold uncertainty score1.000

Codex and Gemma teacher scores by category

CategoryCodexGemma
Metaresearch0.0010.000
Meta-epidemiology (narrow)0.0000.000
Meta-epidemiology (broad)0.0000.000
Bibliometrics0.0000.001
Science and technology studies0.0000.000
Scholarly communication0.0000.000
Open science0.0010.001
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 head0.086
GPT teacher head0.249
Teacher spread0.163 · 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