Lines Are Ideas: Visual Creativity Assessment in Hyperbolic Space
Authors: Baoliang Chen, Zihao Lin, Hao Liu, Xiangjie Sui
Organizations: City University of Hong Kong, Kowloon, Hong Kong SAR, China · South China Normal University, Guangzhou, Guangdong, China · Mango TV (MGTV), Changsha, Hunan, China · City University of Macau, Taipa, Macao SAR, China
Assessing human creativity from drawings provides a scalable way to evaluate human creative ability. However, existing approaches either rely on costly expert annotation or treat creativity as a single regression target, overlooking its underlying structure. Psychological theories have long emphasized that creativity emerges from balancing novelty and appropriateness. Motivated by this principle, we recast creativity as structured semantic deviation between paired drawings, where novelty measures the extent of departure from the original base sketch, and appropriateness determines whether such departures remain meaningful. This formulation naturally leads to a hyperbolic representation, in which novelty is quantified by geodesic distance and appropriateness is evaluated conditionally via a learned decision boundary, further regulated by an abstraction-conditioned adaptive constraint. Experiments on AuDrA-Drawings demonstrate state-of-the-art alignment with expert judgments and strong robustness under rater and task shifts. Beyond performance gains, our framework reveals an interpretable radial organization of creative variation, offering a principled geometric foundation for automatic creativity assessment.
Figures & tables
Figure 1: Motivations . (a) Euclidean geometry is struggles to model the non-uniform, hierarchical concept spaces; (b) hyperbolic geometry provides structure-aware representations via radius- and geodesic-based deviation.
Figure 2: Hyperbolic-space creativity assessment framework .The pipeline consists of four stages: (1) multimodal feature extraction using CLIP encoders to obtain visual features (Fv,Fv) and semantic anchors Ft ; (2) semantic category alignment via cosine similarity to produce content logits zc and abstraction level alignment via an MLP head to obtain zs ; (3) constrained hyperbolic representation by exponential mapping into a Poincaré ball with an abstraction-conditioned adaptive constraint; and (4) creativity scoring via novelty Dnov=dH(h,h) and appropriateness computed from (Dnov,Hsem) , yielding the final creativity score.
Subset
Type
#Samples
Raters
ICC
PD
Abstract
11,075
50
>0.89
RG1
Abstract
670
3 new raters
0.73
RG2
Abstract
722
6 new raters
0.90
FG
Objects
679
3 new raters
0.63
Table 1: Summary of the AuDrA dataset [ 19 ] .
Group
Model
Year
SRCC
PLCC
VGG
VGG16
2014
0.52
0.04
ResNet
ResNet34
2016
0.79
0.81
ResNet50
2016
0.78
0.79
CLIPIQA+
CLIP (ViT-B/32)
2023
0.63
0.65
CLIP (ViT-B/16)
2023
0.67
0.68
CLIP (ViT-L/14)
2023
0.67
0.69
Table 2: Performance comparison on the primary test set.
Group
Model
Year
RG1
RG2
FG
SRCC
PLCC
SRCC
PLCC
SRCC
PLCC
VGG
VGG16
2014
0.58
0.56
0.39
0.32
0.21
0.12
ResNet
ResNet34
2016
0.78
0.75
0.68
0.66
0.45
0.47
ResNet50
2016
0.80
0.77
0.67
0.65
0.46
0.47
CLIPIQA+
CLIP (ViT-B/32)
2023
0.69
0.65
0.41
0.43
0.43
0.44
CLIP (ViT-B/16)
2023
0.73
0.70
0.46
0.47
0.37
0.39
Table 3: Generalization performance on the three unseen subsets in terms of SRCC and PLCC.
Nov.
App.
Adaptive Radius
Hyperbolic
PD
RG1
FG
SRCC / PLCC
SRCC / PLCC
SRCC / PLCC
✓
✓
✓
✓
0.87/0.88
0.79/0.77
0.55/0.56
✗
✓
✓
✓
−0.01/−0.01
−0.06/−0.06
0.05/0.05
✓
✗
✓
✓
0.86/0.85
0.78/0.76
−0.25/−0.20
✓
✓
✗
✓
0.84/0.85
0.78/0.76
0.49/0.51
✓
✓
✗
✗
0.86/0.86
0.78/0.76
0.50/0.50
Table 4: Ablation study results on PD, RG1, and FG. “Nov.” and “App.” denote novelty and appropriateness, respectively.
Figure 3: Correlation between creativity, novelty, and conditional appropriateness. (a) and (b): Correlations of novelty and appropriateness. (b) and (d): Example drawings illustrating how creativity and appropriateness tend to increase with novelty. Under similar novelty levels, higher conditional appropriateness leads to substantially higher creativity scores. “App.” denotes the appropriateness.
Figure 4: Representation magnitude versus abstraction level. (a) Mean norm of CLIP visual embeddings for drawings grouped by abstraction level. (b) Example drawings corresponding to each abstraction category.
Figure 5: Feature geometry in Euclidean vs. hyperbolic space. (a) Euclidean features (b) corresponding hyperbolic embeddings.