task_id
stringlengths 36
64
| origin
imagewidth (px) 1.02k
1.02k
| ours
imagewidth (px) 1.04k
1.04k
| leader
imagewidth (px) 1.04k
1.04k
| winner
stringclasses 3
values | ours_penalty
float64 3
8
| leader_penalty
float64 3
6
| gen_time
float64 19.9
31.4
| ours_ply_size_mb
float64 1.95
83.5
| issues
stringlengths 73
373
|
|---|---|---|---|---|---|---|---|---|---|
0034f67a5a3e1fa3a785dcf12f0e4806de3665879c3b7adb45c02e95e46e3616
|
leader
| 7
| 3
| 22.241306
| 7.59
|
First model is a violin-like instrument with poor detail and wrong shape; second model closely matches the prompt with similar wood texture, beaded edge, and overall form
|
|||
02449b777d8325a7f7e2042b2ee32ba0d1b4f49a09ffa1b1ec9f2cbd0067ae93
|
draw
| 6
| 6
| 26.800225
| 50.67
|
First model has similar flower type but lacks background context and has simplified structure. Second model has different flower structure, added pots, and altered color tones. Both have moderate to major issues with prompt adherence.
|
|||
03bf36122dc1da98a0fa45008c539a4899eec2cf7b659aff5c50fbbf9193d1d2
|
draw
| 3
| 3
| 26.81935
| 49.65
|
Both models have minor shape differences and slight variations in texture/finish compared to the prompt, but overall match the general style and elements (chair with throw and pillow) of the prompt.
|
|||
0581411a50cf5d7f30ab67e9d53ebbe793aadf3b2441c3ca402325326a3db86f
|
draw
| 5
| 4
| 24.386231
| 20.53
|
Both models have significant issues with the flower structure, petal shapes, and mouth details. The first model has jagged petals and mismatched mouth shapes, while the second has different mouth designs and petal textures. Neither matches the prompt's flower structure and mouth details precisely.
|
|||
08716ba11520bf19be5fe52ab460dc5eebb88cc1fc6ec6ae467d6f15f31b6012
|
draw
| 6
| 5
| 24.065578
| 19.36
|
Both models have some shape differences from the prompt, with box details and horn positioning not matching exactly. The first model has boxes with different textures and horn details, while the second has different horn shapes and box decorations. Both have moderate deviations from the prompt's specific design elements.
|
|||
08fca7169c278423c67ecaca8a226c7b5ea4d33d78ba7f81d50a322ed96df312
|
leader
| 6
| 4
| 23.049256
| 62.53
|
First model has major shape and detail issues, second has moderate issues
|
|||
0d1c433bf80c45138f3d795b61a08300b63590a6305b65ac945d46df70e80528
|
draw
| 6
| 6
| 21.392317
| 20.16
|
First model has bowl but missing detailed dragonfly; second model has dragonfly but missing bowl context
|
|||
0e9b8aa1b7c508a02a4a8427fb5d1808fc20978c0ccfc0565a42548f196f5031
|
draw
| 3
| 3
| 23.952627
| 50.93
|
Both models have minor shape differences in the pot and food item compared to the prompt. The pot's color and shape are close but not perfect, and the food item has slight variations in texture and positioning.
|
|||
11a138bc419e5e61e904389c6bb5bcb9fede49a7873ad37316667f96abc9212e
|
leader
| 5
| 3
| 22.534069
| 35.34
|
First model has poor color matching and detail quality, second model has better color but still lacks some texture details
|
|||
11a5f80d11d5fa36989b603419b6135b215cbb1d4d0966942652efae9b104143
|
leader
| 6
| 3
| 21.062893
| 19.06
|
First model has poor rendering quality with transparency issues and lacks detailed elements. Second model has better structure but still misses some details from the prompt.
|
|||
11d30908bf1c6b65318373412b8a0a345cbbb1635dd53e523e245ca93a6ae08b
|
leader
| 5
| 3
| 22.634359
| 40.45
|
First model has minor shape differences and lacks the detailed internal pattern of the prompt; second model has better shape match but still lacks the specific internal pattern detail of the prompt
|
|||
124417db16e4b1b349178621cc957bdc1983431cb019b44ae7d8515a7f6242f0
|
draw
| 5
| 5
| 30.045211
| 10.65
|
First model has some shape differences and missing texture details; second model has major structural and style issues with the hanging mechanism and overall form
|
|||
14f80504c56effdd9bca0a1b6e5227c0083bde89fc874d5c45dc3390cf01e3ce
|
leader
| 5
| 3
| 24.040995
| 24.07
|
First model has wrong shape and missing network pattern details; second model has closer shape but still lacks precise network pattern details
|
|||
16ade48e2ea97d78f97543e81c65f98f0190889163344d82ea6f2341d081561b
|
leader
| 5
| 3
| 19.903752
| 39.29
|
First model has fragmented appearance with multiple panels, while second model has more cohesive stained glass structure. Both maintain forest theme but first has more artistic fragmentation issues, second has better adherence to stained glass form.
|
|||
16c601301eb9677e98f04de0478b3d69e1ada51574cfd51d7e86029a3f81936e
|
leader
| 6.5
| 3
| 24.210949
| 12.14
|
First model has poor texture and shape accuracy, second model has better shape but still lacks detail precision
|
|||
1855c5592cfc3b8556c9c67a2c216bb7fd83826b8333520818ad774611872f49
|
draw
| 4
| 5
| 25.320074
| 61.1
|
First model has minor shape differences and missing small details; second model has more significant style and detail issues
|
|||
1b000364de44fc77208d95f6d155799f86a0def3cf6fee08a5fd4d985971c07d
|
leader
| 7.5
| 3
| 22.136967
| 25.05
|
First model has artistic/watercolor style with distorted proportions and missing detailed features; second model has better shape but still lacks the specific lighting and detail of the prompt image
|
|||
1c548e3ed10a49f79e5df838136ae2a28b503a0d4c5bb3c71fedba0fb197f087
|
draw
| 4
| 3
| 24.119007
| 21.03
|
Both models have minor shape differences and inconsistent bow placement compared to the prompt, but overall match the general top hat concept
|
|||
1e29dc9e121266792cac3d62f1d6353fac10fea2ad57215d677d86c11f813615
|
leader
| 8
| 4
| 21.14235
| 26.95
|
First model has major issues with detail and realism, second model has minor issues with some details
|
|||
1f310cbf91cae354910c44c8a3ae402f010f222e8961e3a89fc0e84582128853
|
draw
| 4
| 3
| 24.434721
| 26.1
|
Both models have minor shape differences and slightly different details compared to the prompt image. The first model has some issues with the roof structure and wheel design, while the second model has similar issues with the roof and some details. Both models generally match the prompt but have minor deviations.
|
|||
1fde24c1fa597076c7ac343480ef30010c7306223f229276615c802270b839ed
|
draw
| 5
| 5
| 25.809197
| 39.78
|
Both models have moderate issues with matching the prompt. They generally resemble a stand mixer but have noticeable differences in shape, detail accuracy, and text elements. Bowl reflections and overall proportions are not perfectly matched to the reference image.
|
|||
203a8159ddbeab2a86a2a819f1b7c2ea24adb90e08ea89db63370eddc92947e3
|
leader
| 5
| 3
| 27.139417
| 46.38
|
First model has black marks, second has better shape but first has more issues
|
|||
2229cf5f-955c-4bc2-91c3-df9255a1c01d
|
leader
| 7.5
| 3
| 23.992323
| 15.1
|
First model has poor rendering quality, wrong gem shape, missing gold details and gem clarity. Second model has better quality but gem shape differs, missing some gold details.
|
|||
22de4efc4723f624b92889e8c79c9b4fb903e8a6b5907c9f0727ede8f2ccab47
|
leader
| 6
| 4
| 22.147909
| 33.21
|
First model has distorted panel shapes and less defined texture; second model has closer panel structure but still some distortion
|
|||
22f497b84d9d8a3426bac9b7496af80a9c31deb0fda9c10c9e9833dea507ad1c
|
draw
| 6
| 5
| 23.820166
| 48.02
|
Both models have moderate issues. They generally match the teal bookshelf structure but have significant differences in book arrangement, object details, and overall style compared to the prompt image. Some shelves and objects are missing or altered, and the overall aesthetic doesn't fully match the prompt's detailed arrangement.
|
|||
24553772-96b9-4c11-a880-56d125634a7a
|
draw
| 5
| 4
| 21.58802
| 32.89
|
Both models have moderate issues with style and detail matching the original prompt. The first model has some shape differences and the second has significant detail variations, though both maintain the general structure of the original building.
|
|||
282c299a1c43900c834d6fac46725fd58812ce8db538df628752eb8452ae0a00
|
draw
| 5
| 4
| 24.546945
| 26.31
|
Both models have moderate issues with style and detail accuracy compared to the prompt. They lack the precision of the original's mechanical details and text markings, with some shape variations and simplified textures.
|
|||
2ce24d6b8dd9f244bfd4c0851dd8f03184d93b0dc7bafc22feabb8c8adef523b
|
leader
| 5
| 3
| 25.516046
| 42.21
|
First model has simplified, less detailed weaving and smoother surfaces; second model has better detail matching the prompt but some proportions differ
|
|||
2ed72b6b9c9b345b17eca522a2584c99018da0f03990817ec082e8babe26291f
|
leader
| 5
| 3
| 22.625764
| 12.32
|
First model has significant detail differences (missing components, different antenna design), second model has closer match but still has some stylistic differences
|
|||
33532e138e69ddba7c5e762eed4bedb6b3fcd1a55302cd784c823d8d76d705f2
|
leader
| 6
| 3
| 23.155981
| 10.93
|
First model has poor texture and detail quality, second model has better detail but some structural differences
|
|||
33827979d7e59540146de3bd1d75e2e1bee4d532e1a61f2426689a0f37baa943
|
draw
| 5
| 4
| 24.058867
| 21.53
|
Both models have moderate issues with shape accuracy and detail matching the original prompt. The first model has some shape differences and the second has style variations that don't fully match the original's delicate, translucent structure.
|
|||
364f2e60a97f04ff4292f98b65e765c5bfbfd61615da2ff56ff9fbad4380f625
|
draw
| 6
| 5
| 24.664457
| 21.88
|
Both models have moderate issues. They generally match the fried banana shape but lack the detailed texture, color consistency, and proper arrangement of the original image. Some models show transparency or incorrect lighting effects.
|
|||
37ca201e96add2bac00fa0ee34a36e0643f287da1e46eb336d6df4f42c708610
|
leader
| 5
| 3
| 23.020687
| 7.71
|
First model has wrong color (dark gray instead of silver), missing handle details, and incorrect texture. Second model has closer color match but still has some shape and handle details issues.
|
|||
38987a28c7b616879a23f7e08ac732df474b9e90684393521f765c04ed0b0fe6
|
draw
| 5
| 4
| 24.136681
| 20.32
|
Both models have some shape differences from the prompt, with the first model having extra handles and the second having a different spout shape. Both lack the detailed texture and aged appearance of the original copper pot, and the background elements are missing.
|
|||
3a0d4173e8316a33298c124eacbba9b20ca3cfed5f6307ee3bc945ab644ab1e1
|
leader
| 6
| 3
| 23.354784
| 12.04
|
First model has wrong color scheme, missing detailed patterns, and incorrect handle shape. Second model has closer color match but still has some shape differences and missing details compared to prompt.
|
|||
3aff357f2c9bc317c083c24d896cdc32f43dd78f0831c3a359f67d08771d23de
|
draw
| 4
| 4
| 25.335914
| 54.12
|
First model has minor shape differences and lighting variations; second model has more significant structural differences and style deviations
|
|||
3df9aac28adfa792e385c45498f7464601e572aaeeaf965ecd96e327f26569e2
|
leader
| 6
| 4
| 30.429737
| 9.04
|
First model has poor texture quality, missing detailed inscriptions, and incorrect color tone. Second model has better detail but still lacks some intricate texturing and precise shape matching.
|
|||
3e743f77623f822d1f1f0264f823576aa8a20bd7344ba0dce92ef8ac63a82547
|
draw
| 6
| 5
| 24.115418
| 57.36
|
Both models have some deviations from the prompt. First model has different glass clarity, tree details, and base texture. Second model has different crown style, glass texture, and tree decorations. Both lack the sharp detail and realistic glass appearance of the prompt image.
|
|||
3f48598320210ac8659a36fd57bccddba1bc5acbab93ac579a0587321088aa84
|
draw
| 4
| 4
| 24.2772
| 26.42
|
First model has minor shape differences and detail variations; second model has more significant style and detail issues, with less accurate floral engravings and different base appearance
|
|||
4024d0d716bab965da116c926ce66d038effa6491958e34b61aa503fdf91ee59
|
leader
| 5
| 3
| 24.071863
| 24.17
|
First model has minor detail issues (missing decorative elements, less polished finish), second model has closer match but still lacks some decorative details compared to prompt
|
|||
4265649c1aaeccc53febeeba273dd30e98a57271e3db940e0ac5652e3e33bd30
|
leader
| 5
| 3
| 22.724557
| 8.21
|
First model has wrong style/color (solid black instead of zebra pattern), second model has partial match with zebra pattern but still has issues with details and style consistency
|
|||
449de7f2e694785e0879713b2b4ce3330941e7bd607d8ff0d67ac5b326aacee8
|
draw
| 5
| 5
| 23.560153
| 49.09
|
First model has minor shape differences and missing small decorative details; second model has more significant style and detail issues, with different pipe arrangement and keyboard appearance
|
|||
44ff32bf-1a77-4e7e-b45a-d8778393748a
|
draw
| 5
| 4
| 28.93716
| 2.79
|
Both models have moderate issues with detail accuracy and style consistency compared to the prompt. They lack the intricate engravings and ornate design elements visible in the original prompt image. The first model has simpler details and less texture variation, while the second model has some improved details but still doesn't match the prompt's level of craftsmanship.
|
|||
4593ab1c-a4fa-48a7-93c0-f17c57fa2af5
|
leader
| 6
| 3
| 23.717654
| 19.89
|
First model has black artifacts and simplified structure; second model has better structure but still has minor shape differences
|
|||
48fb4b27684ae9b4af97387ad0379a4312538bfee3af6ed3253eecb6a5b93d4f
|
leader
| 6
| 4
| 21.728799
| 31.92
|
First model has incorrect shape with open structure, second model has better shape but still lacks precision in faceting details compared to prompt
|
|||
4a46caa383a05ffe10f5ce524096cbd8c0c9916fc621f36a9731583d637846c4
|
leader
| 6
| 3
| 21.654623
| 32.71
|
First model has transparency issues and simplified details; second model has closer match but some detail differences
|
|||
4e9106249bc34ce0ebc0429b6329293a3d9177332f7ba2e9cae930727289cac5
|
draw
| 4
| 3
| 23.851275
| 14.65
|
Both models have similar overall shape but show minor differences in detail placement and texture precision compared to the prompt. The wooden texture and overall form are recognizable but not perfect matches.
|
|||
4f0bdbaa6affcb59438400cfd2d8be5f661e51f313673e4f44205e15a5c0165a
|
draw
| 4
| 3
| 21.118558
| 26.18
|
Both models generally match the prompt but have slight variations in texture, lighting, and shape details. Some have different shading patterns and subtle shape differences from the original prompt.
|
|||
525fbbdb868e86a55d2cae129d03ba345392e8ffdbe7ba6ca99ff1ecdc3cb984
|
draw
| 4
| 4
| 23.6422
| 13.66
|
First model has minor shape differences and background issues; second model has moderate style and detail differences
|
|||
55db1e0e8eb100747a50611d1897d3a8fc4d05e8bc6a2bd3586a43cdcb1a8616
|
draw
| 5
| 4
| 22.882076
| 9.69
|
Both models have moderate issues with style and details. The first model has some shape differences and missing small details, while the second model has wrong style elements and significant details missing compared to the prompt.
|
|||
5806f9f3a8e69b8336a61f98fc0327e98749e8f50cfc00ac6ea7427e2db5cf8b
|
leader
| 6
| 4
| 23.711856
| 17.48
|
First model has poor lighting and detail quality, second model has better detail but lacks laser beams from prompt
|
|||
582b5fa7c5b21b3baf09bf58599e7b4082d21874be3afb3beb6a936a26cc0ffc
|
draw
| 5
| 5
| 22.962141
| 9.9
|
Both models have moderate issues with texture detail, color accuracy, and pose fidelity compared to the prompt image. They lack the fine scale details, specific color gradients, and the naturalistic lighting/reflection seen in the reference image.
|
|||
5a5fbe25dd2291f41fcd4a1e2a2f066cf7d7475edcc602650ecff4bfadbc8028
|
leader
| 5
| 3
| 23.771292
| 12.57
|
First model has wrong style (too smooth, no fabric texture), second model has better shape but still lacks fabric details and background context
|
|||
5aa3a22ae7830606ad5038ea1c7273995d7cd534f883b98fd3402e526ab79391
|
draw
| 6
| 6
| 24.735855
| 33.05
|
First model has distorted shape and feather positioning; second model has simplified structure and missing details
|
|||
5eaf242b81d934958f083ecc80e5eb718e47760646755c0180e640b1401f14f8
|
leader
| 6
| 4
| 24.177611
| 30.46
|
First model has significant visual quality issues with low detail and poor rendering. Second model has better detail but some elements don't match the prompt's specific flower types and arrangement.
|
|||
69847a5d5949d8ce2877b864e7b9ad32ffc7e90b546414d5b79572f0a6669cad
|
draw
| 5
| 4
| 23.747243
| 14.23
|
Both models have similar star-like shapes with pink color and white dots, but they lack the detailed texture and lighting quality of the original prompt image. The first model has less defined details and the second has different shading and texture.
|
|||
6b042c50676b622a978dca8a732a26f8d56a35c8fbf74a4d4a4fcaa7afd886dc
|
draw
| 6
| 6
| 31.391798
| 18.26
|
First model has text variations and inconsistent details; second model has major text and structural issues
|
|||
6e1fa873a0f08a0e1079b858ed64cc40c9dee2c96833edb07fb1522a976dd2e6
|
draw
| 4
| 3
| 23.561522
| 14.66
|
Both models generally match the prompt but have minor shape differences and some detail variations. The teapot shape is recognizable but not identical in proportions and details.
|
|||
70010d16816f40ee956f3e2c396e03fadb5f4982d6c9356f35573d068de44611
|
leader
| 6
| 4
| 24.93038
| 35.85
|
First model has poor texture detail and shape accuracy; second model has better shape but still lacks fine texture details compared to prompt
|
|||
7171ad88e62858e96cbcd9f6cf1cad5c89f1bfb818f5d208ceacd24e489adb34
|
leader
| 6
| 4
| 24.189874
| 14.15
|
First model has poor texture and detail, second model has better shape but still lacks fine details and realistic pearl appearance
|
|||
71bc1ac0c0eb9bd1be4a7150a602ad7727947ed1c5248aba711d47690ea14e3a
|
draw
| 5
| 6
| 24.281705
| 81.07
|
First model has some detail differences (crystal arrangement, box structure), second model has major structural and decorative deviations from original prompt
|
|||
7811aa32454d6f1cf9ef5ff2a6958d94140e20c33c2fae079debfd26809c3539
|
draw
| 4
| 3
| 30.217476
| 15.44
|
Both models have similar shape but differ in lighting/shading details and base appearance. Some minor details like bottle texture and base design don't match the original prompt exactly.
|
|||
78eef3555870d070d454d48893f45d9d1d932da566a4c0bfb4701037d761eaed
|
draw
| 6
| 5
| 22.681828
| 7.7
|
Both models have moderate issues with style and details. They lack the detailed texture, lighting, and specific design elements (like the cockpit window shape, panel details, and overall weathering) of the original prompt image. The first model has more shape differences, while the second has different proportions and missing details.
|
|||
7976c38065b686a9d704de5b957e2acddf12c6a77979713f245262354ae776f2
|
leader
| 5
| 3
| 25.622253
| 43.49
|
First model has moderate issues with texture and detail accuracy, second has minor issues with overall shape and detail matching
|
|||
79dba7120244f263228580c4069f51578a30f97304cc2f93c7b7c0666e9eb6b6
|
draw
| 5
| 5
| 25.227154
| 36.27
|
First model has some detail variations (e.g. unicorn features, floral patterns), second model has major structural and detail deviations from original prompt
|
|||
7c7df949fd112c9a3b33311bcfa387160927231a5aa41cb3a0064618d1abe87e
|
draw
| 6
| 6
| 27.857017
| 59.28
|
First model has some shape differences and book arrangement issues; second model has major structural and style deviations
|
|||
854c1449-39a1-4846-8af4-d6ebe8d4f619
|
draw
| 4
| 4
| 24.166295
| 25.85
|
First model has some shape differences and missing small details; second model has more significant issues with style and details
|
|||
8916afced891fe7c446c3e06d522c9ddff2605fffff84b578ba0dea1f4495730
|
leader
| 5
| 3
| 23.980595
| 25.32
|
First model has black elements and different structural details; second model has closer match but still has some detail differences
|
|||
89e1b8998cc50f9be2f55127b09effa9030e523bec6d65b72f74d4c4db06099b
|
draw
| 4
| 4
| 24.654331
| 27.51
|
First model has minor shape differences and slight detail variations; second model has more noticeable style and detail differences
|
|||
8a08215ae9ee2976fbddabdca5bc5046c9c8b9dc5829a336cf3bcba249fc00f5
|
leader
| 5
| 3
| 24.863442
| 29.59
|
First model has poor detail and transparency issues; second model has better detail but some stylistic differences
|
|||
8dbd97f96c4d02a66fac4a71ae0d6377459b3400dca489574454a5bbecb57154
|
draw
| 4
| 4
| 21.869955
| 24.2
|
First model has minor shape differences and some detail variations; second model has more significant style and detail issues, with less accurate representation of the original's texture and decoration placement
|
|||
92036f3abc8c73b3d2185d7afec12ab838bf0d50910b31c0c1d4f4114442d004
|
ours
| 4
| 6
| 21.998691
| 35.9
|
First model has some shape differences and missing environmental details; second model has major architectural style deviations and missing environmental elements
|
|||
92b6f4511409351316f198024d3b849b9d86d469a6549722e70e0c0cee504916
|
leader
| 5
| 3
| 20.822686
| 21.67
|
First model has significant texture and detail issues, second model is closer but still has some detail differences
|
|||
9878bf40b07cfc675cc42f91cfd4d911c3856c8e2258e19f7c0f7536320c0b64
|
draw
| 6
| 5
| 24.793933
| 37.66
|
Both models have moderate issues. They generally match the structure but have significant detail differences in book arrangement, decorative elements, and overall style. Some elements like the central decorative piece and book colors don't match the prompt as closely as needed.
|
|||
988209beaf8802e7b01184b774fdbbc2b17bcac96dd102343e6bad1527504ee9
|
draw
| 3
| 3
| 23.173187
| 16.28
|
Both models generally match the prompt but have minor issues. The basket texture and flower arrangement have slight variations from the original image. Some details like basket weave pattern and flower positioning aren't perfectly matched.
|
|||
9892e6f5c8dc0ccb7b453f21860dc2be3f7be0157c5add3c1943c65e1b278f44
|
leader
| 5
| 3
| 23.056528
| 47.45
|
First model has lower detail and less accurate shape matching, second model has closer detail but still some differences in proportions and specific features
|
|||
9923a943b8a628ee3a16ca4111f5010e876e90b4f491e2cbd8a036f0bd4cb750
|
draw
| 4
| 3
| 30.209045
| 9.42
|
Slight shape differences and minor detail variations in the 3D models compared to the prompt image
|
|||
99b7192fd902db75f393fda95ce3f811278636a6df92f6c691dc819890771116
|
leader
| 6
| 3
| 24.056248
| 21.84
|
First model has semi-transparent elements and simplified structure; second model has more accurate metal framework but still has some detail differences
|
|||
9b22fb789fbc3d9e7cd50e4cc21300bcfac60930735d4653d97dfd6d9316ef84
|
leader
| 5
| 3
| 29.666498
| 11.36
|
First model has minor shape differences and some detail issues; second model has closer match but still has some detail differences
|
|||
9c515261c35590822a4ff6f1fe5781aa55da3453bb8be5904aa2f0b1bf0fd0a8
|
draw
| 5
| 4
| 21.331814
| 25.99
|
Both models have some shape differences from the prompt, with minor detail issues and style variations. The first model has different lighting and color tones, while the second has some structural differences in the lantern design.
|
|||
9c6df3b2c14e9a312d3236417e1aea3aab531b70c487edfd1eeeee2fabbc79d2
|
draw
| 5
| 4
| 23.86354
| 20.53
|
Both models have moderate issues with style and detail matching the prompt. The first model has some shape differences and the second has significant detail variations, though both maintain the general L-shape and color scheme.
|
|||
9dbc68f175b56e60ae95f9511fb94546bbf4b49c0f5b7e5bf5b607cb38b19bc1
|
draw
| 6
| 5
| 29.362092
| 7.79
|
Both models have some shape differences from the prompt, with less realistic glass texture, altered proportions, and simplified details. The rope/knot structure is less detailed, and the overall lighting/texture quality doesn't match the prompt's photorealistic style.
|
|||
a4d44ab3b65677ae3e2194f4d9fe32a7d76ae039d5fc6fb684dd78a2fbf0adf9
|
draw
| 5
| 4
| 21.870481
| 32.23
|
Both models have similar but not perfect matches. They have the general shape of a beer glass with foam, but there are differences in glass texture, foam consistency, and lighting. Some details like the exact glass shape and foam texture don't match the reference image precisely.
|
|||
a60c5b2c3dafb223d0ecbd408995e8bb346960822148f78b30c106b1375e1ad6
|
leader
| 6
| 3
| 24.435176
| 21.23
|
First model has artistic, abstract rendering with poor detail matching; second model has better detail but still lacks background field context
|
|||
a77062a2d0222debdbd68d33e96865cda4df9ac6a18d9f0a507033a9c79002e0
|
draw
| 5
| 4
| 24.096264
| 25.39
|
Both models have some shape differences from the prompt, with minor detail issues and some style variations. The eagle's position, ship details, and overall aesthetic don't perfectly match the prompt's wooden texture and specific design elements.
|
|||
aacb989d16471a364824ac145d8b06fa273e3a8ac0af3ed6ea78e1656777d830
|
draw
| 4
| 3
| 23.913227
| 24.46
|
Both models have minor shape differences and some missing small details compared to the prompt, but overall match the DeLorean style and general features
|
|||
ab455629abc3223b6b448253da59120dccdceccf123354ceadd5a904e2bed8d3
|
leader
| 5
| 3
| 22.514415
| 35.09
|
First model has simplified features and less detailed drapery; second model has closer resemblance but still lacks some fine details of the original prompt's sculpture
|
|||
ab506d408112974e5ba8f2a6dcae249e7c2674b6e60f0dff4053a9ddc28fbd33
|
draw
| 5
| 4
| 21.437758
| 22.58
|
Both models have similar shape but differ in flower details, color intensity, and pattern consistency with the prompt. Some flowers appear less defined, and color matching isn't perfect.
|
|||
accbba278390739a94975fe2a61c53e7b30e6fa17cb8e8b9bb7bc572a5fb8d44
|
draw
| 5
| 5
| 26.512599
| 42.71
|
First model has some shape differences and missing detailed textures; second model has major structural changes with circular elements not in prompt
|
|||
acd3d46a7d9b6cfbc7337bbfc6a785524933994f198df4d4c6211721d33846c6
|
draw
| 4
| 3
| 23.886552
| 14.15
|
Both models have minor shape differences and some details don't match exactly, but overall structure and basic features align with the prompt
|
|||
aebc379c6b12425cd01b6ddd3f6adc644c2ad09b27300f48f2d9f7d1ccb65f8a
|
draw
| 6
| 5
| 23.404858
| 12.53
|
Both models have some shape differences from the original prompt, with variations in candle counts, flame appearance, and candelabra structure. Details like the gold texture and overall proportions are not perfectly matched.
|
|||
afd25f32172975ac089bf170a52a6fea54fd6270042ad3c392fcd5d16dd1da1d
|
draw
| 6
| 5
| 30.071151
| 10.84
|
Both models have moderate issues with style and details. The first model has more varied textures and lighting issues, while the second has different proportions and base details. Neither fully matches the original prompt's level of detail and style.
|
|||
b110ed3a3548f903076d10ed37efd1117e153cb0248215db63c8c27eb31f0144
|
draw
| 5
| 4
| 21.3846
| 18.84
|
Both models have color schemes similar to the prompt but lack the detailed patchwork texture and specific shape precision of the original. The first model has softer colors and less defined patches, while the second has different patch arrangements and less accurate color matching.
|
|||
b57421b571110de4bc492bce2481407f7740fc4242178a232f0dc34246bc1df4
|
draw
| 5
| 4
| 23.426979
| 13.76
|
Both models have rainbow manes/tails but lack the puzzle piece structure and lighting quality of the original prompt. First model has transparency issues and second has different proportions.
|
|||
b8a0a728fa4a2f2b7160fa0f5a8df2dbb0d53b338010ce310d1d3e1625d1de92
|
leader
| 5
| 3
| 23.503853
| 9.93
|
First model has wrong style (not matching the sleek futuristic design), second model has closer match but still lacks the neon lighting and futuristic environment details
|
|||
ba495014c7920c09602f1ecc0c75c051f005dcb6a8bbf132d07ede2ee43b9e32
|
leader
| 5
| 3
| 23.650212
| 22.41
|
First model has poor texture quality and missing detailed patterns; second model has better detail but some proportions differ
|
|||
bce2d1db387cfd5acf13509c542fe3cd722b31a11980419487b502f830cd4a44
|
leader
| 5
| 3
| 24.970141
| 35.45
|
First model has issues with texture quality and some details, second model has closer match but still has some detail differences
|
|||
bde166eae9a467e70699be378dfd90814e7fa9637899667631fd2729fb7bdf47
|
draw
| 3
| 3
| 22.640404
| 6.69
|
Both models have similar overall structure but show minor shape differences in body proportions, head details, and some mechanical elements. Details like the head shape, body contours, and specific mechanical joints differ slightly from the prompt image.
|
|||
be94887e8e032e46d4bbd2f2024382c94b103450d42bf7b9ce9eb28198be507c
|
leader
| 5
| 3
| 23.429192
| 13.24
|
First model has lower detail and some shape inaccuracies; second model has closer match but still lacks some details compared to prompt
|
|||
c1abfbc5fb9bc7f6d31a03986d730328e1757e24a12f27824fc77ef17ff51578
|
draw
| 6
| 5
| 27.797486
| 54.45
|
Both models have moderate issues. The first model has simplified crystal shapes and less detail, while the second has different crystal arrangements and less texture. Both lack the detailed, realistic appearance of the original prompt.
|
End of preview. Expand
in Data Studio
YAML Metadata
Warning:
empty or missing yaml metadata in repo card
(https://huggingface.co/docs/hub/datasets-cards)
Benchmark Results: v2.3.0
Overview
- Round: 14
- Test Name: v2.3.0
- Timestamp: 2026-01-16 18:34:22
Summary
| Metric | Value |
|---|---|
| Total Tasks | 128 |
| Total Judged | 128 |
| Skipped | 0 |
| Gen Failed | 0 |
| Render Failed | 0 |
| Judge Failed | 0 |
Results
| Result | Count |
|---|---|
| 🏆 WINS | 1 |
| 🤝 DRAWS | 74 |
| ❌ LOSSES | 53 |
Performance
| Metric | Value |
|---|---|
| Win Rate | 0.8% |
| Margin | -40.6% |
| Avg Gen Time | 24.11s |
Methodology
- Judge Model: thudm/glm-4.1v-9b-thinking (via OpenRouter)
- Position-Balanced: 2 passes with swapped order to avoid position bias
- Penalty Scale: 0 (perfect) to 10 (completely wrong)
- Winner: Lower average penalty wins (draw if difference ≤ 1)
Files
benchmark_v2.3.0.json- Full benchmark results*/ours.png- Our rendered 3D models*/leader.png- Leader's rendered 3D models*/origin.png- Original prompt images
- Downloads last month
- 29