AI Paper Atlas

Research map / Video / Video segmentation and tracking

GeoStyle: Discovering Fashion Trends and Events

Utkarsh Mall, Kevin Matzen, Bharath Hariharan, Noah Snavely, Kavita Bala

ICCV 2019 · Video › Video segmentation and tracking · approach: Spatio-temporal predictive modeling · 69 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use the full parametric model (linear plus cyclic terms) instead of linear only (ablation)trend model · Attribute trends, next 26 weeksMAE: 0.0365 -> 0.0145Improved
use the full model instead of sinusoid only (ablation)trend model · Attribute trends, next 26 weeksMAE: 0.0163 -> 0.0145Improved
use the full model instead of cyclic only (ablation)trend model · Attribute trends, next 26 weeksMAE: 0.0165 -> 0.0145Improved
use the full model instead of linear+sinusoid (ablation)trend model · Attribute trends, next 26 weeksMAE: 0.0169 -> 0.0145Improved
use the full model instead of cyclic only (ablation)trend model · Attribute trends, next weekMAE: 0.0129 -> 0.0119Improved
use parametric model instead of linear only (ablation)trend model · Style-cluster trends, next 26 weeksMAE: 0.0135 -> 0.0077Improved
use parametric model instead of autoregression (up to 12x more parameters)trend model · Attribute trends, next 26 weeksMAE: 0.0207 -> 0.0145Improved
use parametric model instead of exponential smoothingtrend model · Attribute trends, next 26 weeksMAPE: 20.59 -> 15.73Improved
use parametric model instead of VARtrend model · Style-cluster trends, next 26 weeksMAPE: 27.97 -> 21.78Improved
longer forecast horizon (next week to next 26 weeks) widens gap to baselinestrend model · Attribute trendsMAE: 0.0152 -> 0.0231Improved
add full model on unseen city Barcelona vs sinusoid+linear (ablation)trend model · Barcelona unseen cityMAE: 0.043 -> 0.043No clear effect
use parametric model instead of Autoregression on unseen citytrend model · Barcelona unseen cityMAE: 0.047 -> 0.043Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy