HaruthaiAi commited on
Commit
24f1ce5
Β·
verified Β·
1 Parent(s): 44c058f

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +78 -7
README.md CHANGED
@@ -8,17 +8,88 @@ sdk_version: 5.35.0
8
  app_file: app.py
9
  pinned: false
10
  license: creativeml-openrail-m
11
- short_description: AI analysis comparing Tree Oil Painting and Van Gogh’s Los D
12
  ---
13
 
14
- Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
15
 
16
  ---
17
- ### 🎨 Van Gogh – *Los Descargadores en Arles* (1888)
18
 
19
- This painting, created by Vincent van Gogh in 1888, depicts dock workers at sunset along the RhΓ΄ne River.
20
- The intense yellow sky and bold silhouettes showcase Van Gogh's expressive brushwork and emotional use of color.
21
 
22
- In this Space, we use AI to compare its brushstroke structure and torque rhythm to *The Tree Oil Painting*,
23
- revealing hidden patterns and stylistic parallels between the two works.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
24
 
 
 
8
  app_file: app.py
9
  pinned: false
10
  license: creativeml-openrail-m
11
+ short_description: Tree Oil vs Van Gogh (1888) – AI brushstroke comparison
12
  ---
13
 
14
+ # Tree Oil vs Van Gogh (1888) – AI Brushstroke Analysis
15
+
16
+ This Hugging Face Space presents a comparative AI analysis between **The Tree Oil Painting** and **Vincent van Gogh’s Los Descargadores en Arles (1888)** using a suite of brushstroke forensics called the **18 Supreme Techniques**.
17
+
18
+ The goal is to examine structural parallels in torque, brushstroke rhythm, energy dispersion, and underdrawing across both works.
19
+
20
+ ---
21
+
22
+ ## πŸ”¬ Methodology & Rule
23
+
24
+ All comparisons follow a strict validation protocol:
25
+
26
+ - **SSIM is not used** (prohibited).
27
+ - **AI Natural Matching only** is applied.
28
+ - *The Tree Oil Painting* is used as the **master reference**.
29
+ - The analysis framework includes torque estimation, pigment flow, frequency structure, and vanishing point detection.
30
+
31
+ ---
32
+
33
+ ## πŸ“Š Key Analysis Results
34
+
35
+ ### 1. Tree Oil – Original Painting
36
+ ![Tree Oil](./treeoil_full.jpg)
37
+
38
+ ---
39
+
40
+ ### 2. Tree Oil – X-ray & Underdrawing
41
+ ![Xray Type I](./treeoil_xray_type1.jpg)
42
+ ![Xray Type II](./treeoil_xray_facehighlight.jpg)
43
+
44
+ ---
45
+
46
+ ### 3. Tree Oil – Sobel Energy Consistency
47
+ ![Sobel Consistency](./sobel_energy_consistency.jpg)
48
+
49
+ ---
50
+
51
+ ### 4. Tree Oil – 18 Supreme Techniques Grid
52
+ ![TreeOil 18Tech](./treeoil_sample_18tech_spaceversion_for_vangogh_compare.jpg)
53
 
54
  ---
 
55
 
56
+ ### 5. Tree Oil – Sobel Magnitude Over Time (Old vs New 2025)
57
+ ![Sobel Graph](./sobel_graph_old_vs_new.jpg)
58
 
59
+ ---
60
+
61
+ ## πŸ“ˆ Technique-by-Technique Comparison with Van Gogh (1888)
62
+
63
+ ### 6. Score Comparison – 18 Techniques
64
+ ![Score Bar Chart](./treeoil_vangogh1888_consistency_scores.jpg)
65
+
66
+ ---
67
+
68
+ ## 🎨 Van Gogh’s Los Descargadores en Arles (1888)
69
+
70
+ ### 7. Original Painting
71
+ ![Van Gogh Original](./VanGogh_Los_Descargadores_en_Arles_1888.jpg)
72
+
73
+ ---
74
+
75
+ ### 8. 18 Technique Grid – Van Gogh (1888)
76
+ ![VG 18Tech](./VanGogh_LosDescargadores_1888_18SupremeTech_ResultGrid.jpg)
77
+
78
+ ---
79
+
80
+ ## 🧠 Observations
81
+
82
+ - **Stroke pressure**, **underdrawing**, and **vanishing torque** show extremely high similarity.
83
+ - Sobel energy patterns from 2025 analysis match both old and new models, indicating strong internal consistency.
84
+ - X-ray overlays reveal structured underlayers, possibly symbolic or reused canvas areas.
85
+
86
+ ---
87
+
88
+ ## πŸ“Œ Conclusion
89
+
90
+ This forensic comparison highlights statistically significant AI-based similarities between *The Tree Oil Painting* and Van Gogh’s 1888 work.
91
+ The project is part of a broader open research effort to let AI and the public reevaluate overlooked artworks using transparent, reproducible methods.
92
+
93
+ ---
94
 
95
+ *Created and curated by HaruthaiAi – All findings are open for peer review and further model training.*