AI Paper Atlas

Research map / Detection and segmentation / Open-set and few-shot detection

SPIN: Hierarchical Segmentation with Subpart Granularity in Natural Images

Josh David Myers-Dean, Jarek T Reynolds, Brian Price, Yifei Fan, Danna Gurari

ECCV 2024 · Detection and segmentation › Open-set and few-shot detection · approach: Interactive and referring segmentation · 7 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
fine-tune GLaMM on SPIN training dataGLaMM 7B · SPIN subparts (General prompt)mIoU_S: 11.00 -> 24.56Improved
fine-tune GLaMM on SPIN training dataGLaMM 7B · SPIN parts (General prompt)mIoU_P: 40.00 -> 60.76Improved
fine-tune GLaMM on SPIN training dataGLaMM 7B · SPIN objects (General prompt)mIoU_O: 86.31 -> 91.08Improved
use hierarchical-specific HIPIE instead of general foundation modelshierarchical segmentation · SPIN subpart localizationmIoU_S: 0.92 (HIPIE ViT-H) vs 11.55 (LISA 13B)Hurt
prompt with general rather than specific object categories (ablation)CoGVLM, LISA · SPIN objectsmIoU_O: -2.82pp CoGVLM; -2.33pp LISA-13B; -2.87pp LISA-7BHurt
general versus specific prompts for parts and subparts (ablation)foundation models · SPIN parts and subpartsmIoU: gap within 1ppNo clear effect
give SAM ground-truth boxes instead of language-only prompting (versus GLaMM)SAM · SPIN subpartsmIoU_S: +38.61pp subparts, +20.23pp parts, +3.75pp objectsImproved
pair SAM with CoGVLM predicted boxes versus GLaMMSAM + CoGVLM · SPIN subparts, parts and objectsmIoU: +9.24pp subparts, +10.87pp parts, -11.22pp objectsMixed
move from object to part to subpart granularityall models · SPIN zero-shot localizationmIoU: 86 -> 40 -> 14 (best object, part, subpart)Hurt
scale ViP-Llava from 7B to 13B for recognitionViP-Llava · SPIN recognitionmACC_S (%): 53.57 -> 94.78Improved

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