Dataset Viewer (First 5GB)
Auto-converted to Parquet Duplicate
id
int64
image
list
conversations
list
0
[ "coco2017/train2017/000000000009.jpg", "coco2017/depth/000000000009_depth.png" ]
[ { "from": "human", "value": "<image>Predict relative depth visualized as a grayscale image with subtle, continuous shading transitions. The output should show gradual value changes throughout, not discrete color regions like segmentation masks." }, { "from": "gpt", "value": "<image>" } ]
1
[ "coco2017/train2017/000000000030.jpg", "coco2017/depth/000000000030_depth.png" ]
[ { "from": "human", "value": "<image>Predict relative depth visualized as a grayscale image with subtle, continuous shading transitions. The output should show gradual value changes throughout, not discrete color regions like segmentation masks." }, { "from": "gpt", "value": "<image>" } ]
2
[ "coco2017/train2017/000000000034.jpg", "coco2017/depth/000000000034_depth.png" ]
[ { "from": "human", "value": "<image>Output a grayscale image with continuous tone variations representing depth. Unlike segmentation discrete color blocks, this should show smooth brightness gradients across the entire image." }, { "from": "gpt", "value": "<image>" } ]
3
[ "coco2017/train2017/000000000042.jpg", "coco2017/depth/000000000042_depth.png" ]
[ { "from": "human", "value": "<image>Predict per-pixel relative depth, visualized as a continuous grayscale image where brightness indicates proximity (brighter for closer, darker for farther). The output should be a smooth gray image with varying shades of gray, representing a continuous depth field." }, ...
4
[ "coco2017/train2017/000000000072.jpg", "coco2017/depth/000000000072_depth.png" ]
[ { "from": "human", "value": "<image>Visualize relative depth as a continuous grayscale image. The output should show gradual value transitions without discrete boundaries, distinguishing it from segmentation blocky color regions." }, { "from": "gpt", "value": "<image>" } ]
5
[ "coco2017/train2017/000000000073.jpg", "coco2017/depth/000000000073_depth.png" ]
[ { "from": "human", "value": "<image>Estimate relative depth for each pixel in the image, with closer objects appearing brighter and distant objects appearing darker. Output is a grayscale image with pixel values ranging from 0-255." }, { "from": "gpt", "value": "<image>" } ]
6
[ "coco2017/train2017/000000000074.jpg", "coco2017/depth/000000000074_depth.png" ]
[ { "from": "human", "value": "<image>Predict per-pixel relative depth, visualized as a continuous grayscale image where brightness indicates proximity (brighter for closer, darker for farther). The output should be a smooth gray image with varying shades of gray, representing a continuous depth field." }, ...
7
[ "coco2017/train2017/000000000078.jpg", "coco2017/depth/000000000078_depth.png" ]
[ { "from": "human", "value": "<image>Predict per-pixel relative depth, visualized as a continuous grayscale image where brightness indicates proximity (brighter for closer, darker for farther). The output should be a smooth gray image with varying shades of gray, representing a continuous depth field." }, ...
8
[ "coco2017/train2017/000000000086.jpg", "coco2017/depth/000000000086_depth.png" ]
[ { "from": "human", "value": "<image>Estimate relative depth for each pixel in the image, with closer objects appearing brighter and distant objects appearing darker. Output is a grayscale image with pixel values ranging from 0-255." }, { "from": "gpt", "value": "<image>" } ]
9
[ "coco2017/train2017/000000000089.jpg", "coco2017/depth/000000000089_depth.png" ]
[ { "from": "human", "value": "<image>Estimate relative depth for each pixel in the image, with closer objects appearing brighter and distant objects appearing darker. Output is a grayscale image with pixel values ranging from 0-255." }, { "from": "gpt", "value": "<image>" } ]
10
[ "coco2017/train2017/000000000092.jpg", "coco2017/depth/000000000092_depth.png" ]
[ { "from": "human", "value": "<image>Output a grayscale image with continuous tone variations representing depth. Unlike segmentation discrete color blocks, this should show smooth brightness gradients across the entire image." }, { "from": "gpt", "value": "<image>" } ]
11
[ "coco2017/train2017/000000000110.jpg", "coco2017/depth/000000000110_depth.png" ]
[ { "from": "human", "value": "<image>Estimate relative depth for each pixel in the image, with closer objects appearing brighter and distant objects appearing darker. Output is a grayscale image with pixel values ranging from 0-255." }, { "from": "gpt", "value": "<image>" } ]
12
[ "coco2017/train2017/000000000113.jpg", "coco2017/depth/000000000113_depth.png" ]
[ { "from": "human", "value": "<image>Visualize relative depth as a continuous grayscale image. The output should show gradual value transitions without discrete boundaries, distinguishing it from segmentation blocky color regions." }, { "from": "gpt", "value": "<image>" } ]
13
[ "coco2017/train2017/000000000127.jpg", "coco2017/depth/000000000127_depth.png" ]
[ { "from": "human", "value": "<image>Estimate relative depth per pixel. Output a grayscale image where brighter pixels represent closer distances. The result should show continuous tone variations without discrete color jumps, distinguishing it from other task outputs." }, { "from": "gpt", "value...
14
[ "coco2017/train2017/000000000133.jpg", "coco2017/depth/000000000133_depth.png" ]
[ { "from": "human", "value": "<image>Predict per-pixel relative depth, visualized as a continuous grayscale image where brightness indicates proximity (brighter for closer, darker for farther). The output should be a smooth gray image with varying shades of gray, representing a continuous depth field." }, ...
15
[ "coco2017/train2017/000000000136.jpg", "coco2017/depth/000000000136_depth.png" ]
[ { "from": "human", "value": "<image>Output a grayscale image with continuous tone variations representing depth. Unlike segmentation discrete color blocks, this should show smooth brightness gradients across the entire image." }, { "from": "gpt", "value": "<image>" } ]
16
[ "coco2017/train2017/000000000142.jpg", "coco2017/depth/000000000142_depth.png" ]
[ { "from": "human", "value": "<image>Predict relative depth visualized as a grayscale image with subtle, continuous shading transitions. The output should show gradual value changes throughout, not discrete color regions like segmentation masks." }, { "from": "gpt", "value": "<image>" } ]
17
[ "coco2017/train2017/000000000144.jpg", "coco2017/depth/000000000144_depth.png" ]
[ { "from": "human", "value": "<image>Predict per-pixel relative depth, visualized as a continuous grayscale image where brightness indicates proximity (brighter for closer, darker for farther). The output should be a smooth gray image with varying shades of gray, representing a continuous depth field." }, ...
18
[ "coco2017/train2017/000000000151.jpg", "coco2017/depth/000000000151_depth.png" ]
[ { "from": "human", "value": "<image>Output a grayscale image with continuous tone variations representing depth. Unlike segmentation discrete color blocks, this should show smooth brightness gradients across the entire image." }, { "from": "gpt", "value": "<image>" } ]
19
[ "coco2017/train2017/000000000154.jpg", "coco2017/depth/000000000154_depth.png" ]
[ { "from": "human", "value": "<image>Predict relative depth visualized as a grayscale image with subtle, continuous shading transitions. The output should show gradual value changes throughout, not discrete color regions like segmentation masks." }, { "from": "gpt", "value": "<image>" } ]
20
[ "coco2017/train2017/000000000164.jpg", "coco2017/depth/000000000164_depth.png" ]
[ { "from": "human", "value": "<image>Visualize relative depth as a continuous grayscale image. The output should show gradual value transitions without discrete boundaries, distinguishing it from segmentation blocky color regions." }, { "from": "gpt", "value": "<image>" } ]
21
[ "coco2017/train2017/000000000165.jpg", "coco2017/depth/000000000165_depth.png" ]
[ { "from": "human", "value": "<image>Predict relative depth visualized as a grayscale image with subtle, continuous shading transitions. The output should show gradual value changes throughout, not discrete color regions like segmentation masks." }, { "from": "gpt", "value": "<image>" } ]
22
[ "coco2017/train2017/000000000192.jpg", "coco2017/depth/000000000192_depth.png" ]
[ { "from": "human", "value": "<image>Predict per-pixel relative depth, visualized as a continuous grayscale image where brightness indicates proximity (brighter for closer, darker for farther). The output should be a smooth gray image with varying shades of gray, representing a continuous depth field." }, ...
23
[ "coco2017/train2017/000000000194.jpg", "coco2017/depth/000000000194_depth.png" ]
[ { "from": "human", "value": "<image>Visualize relative depth as a continuous grayscale image. The output should show gradual value transitions without discrete boundaries, distinguishing it from segmentation blocky color regions." }, { "from": "gpt", "value": "<image>" } ]
24
[ "coco2017/train2017/000000000196.jpg", "coco2017/depth/000000000196_depth.png" ]
[ { "from": "human", "value": "<image>Generate a continuous depth map as a grayscale image with smooth variations in gray levels (0-255). Closer regions appear brighter, farther regions appear darker, creating a visually continuous depth representation. Output is a image" }, { "from": "gpt", "valu...
25
[ "coco2017/train2017/000000000201.jpg", "coco2017/depth/000000000201_depth.png" ]
[ { "from": "human", "value": "<image>Generate a continuous depth map as a grayscale image with smooth variations in gray levels (0-255). Closer regions appear brighter, farther regions appear darker, creating a visually continuous depth representation. Output is a image" }, { "from": "gpt", "valu...
26
[ "coco2017/train2017/000000000208.jpg", "coco2017/depth/000000000208_depth.png" ]
[ { "from": "human", "value": "<image>Generate a continuous depth map as a grayscale image with smooth variations in gray levels (0-255). Closer regions appear brighter, farther regions appear darker, creating a visually continuous depth representation. Output is a image" }, { "from": "gpt", "valu...
27
[ "coco2017/train2017/000000000241.jpg", "coco2017/depth/000000000241_depth.png" ]
[ { "from": "human", "value": "<image>Output a grayscale image with continuous tone variations representing depth. Unlike segmentation discrete color blocks, this should show smooth brightness gradients across the entire image." }, { "from": "gpt", "value": "<image>" } ]
28
[ "coco2017/train2017/000000000263.jpg", "coco2017/depth/000000000263_depth.png" ]
[ { "from": "human", "value": "<image>Estimate relative depth for each pixel in the image, with closer objects appearing brighter and distant objects appearing darker. Output is a grayscale image with pixel values ranging from 0-255." }, { "from": "gpt", "value": "<image>" } ]
29
[ "coco2017/train2017/000000000283.jpg", "coco2017/depth/000000000283_depth.png" ]
[ { "from": "human", "value": "<image>Generate a continuous depth map as a grayscale image with smooth variations in gray levels (0-255). Closer regions appear brighter, farther regions appear darker, creating a visually continuous depth representation. Output is a image" }, { "from": "gpt", "valu...
30
[ "coco2017/train2017/000000000294.jpg", "coco2017/depth/000000000294_depth.png" ]
[ { "from": "human", "value": "<image>Generate a continuous depth map as a grayscale image with smooth variations in gray levels (0-255). Closer regions appear brighter, farther regions appear darker, creating a visually continuous depth representation. Output is a image" }, { "from": "gpt", "valu...
31
[ "coco2017/train2017/000000000307.jpg", "coco2017/depth/000000000307_depth.png" ]
[ { "from": "human", "value": "<image>Estimate relative depth per pixel. Output a grayscale image where brighter pixels represent closer distances. The result should show continuous tone variations without discrete color jumps, distinguishing it from other task outputs." }, { "from": "gpt", "value...
32
[ "coco2017/train2017/000000000309.jpg", "coco2017/depth/000000000309_depth.png" ]
[ { "from": "human", "value": "<image>Predict relative depth visualized as a grayscale image with subtle, continuous shading transitions. The output should show gradual value changes throughout, not discrete color regions like segmentation masks." }, { "from": "gpt", "value": "<image>" } ]
33
[ "coco2017/train2017/000000000321.jpg", "coco2017/depth/000000000321_depth.png" ]
[ { "from": "human", "value": "<image>Predict per-pixel relative depth, visualized as a continuous grayscale image where brightness indicates proximity (brighter for closer, darker for farther). The output should be a smooth gray image with varying shades of gray, representing a continuous depth field." }, ...
34
[ "coco2017/train2017/000000000322.jpg", "coco2017/depth/000000000322_depth.png" ]
[ { "from": "human", "value": "<image>Predict per-pixel relative depth, visualized as a continuous grayscale image where brightness indicates proximity (brighter for closer, darker for farther). The output should be a smooth gray image with varying shades of gray, representing a continuous depth field." }, ...
35
[ "coco2017/train2017/000000000326.jpg", "coco2017/depth/000000000326_depth.png" ]
[ { "from": "human", "value": "<image>Estimate relative depth for each pixel in the image, with closer objects appearing brighter and distant objects appearing darker. Output is a grayscale image with pixel values ranging from 0-255." }, { "from": "gpt", "value": "<image>" } ]
36
[ "coco2017/train2017/000000000332.jpg", "coco2017/depth/000000000332_depth.png" ]
[ { "from": "human", "value": "<image>Estimate relative depth for each pixel in the image, with closer objects appearing brighter and distant objects appearing darker. Output is a grayscale image with pixel values ranging from 0-255." }, { "from": "gpt", "value": "<image>" } ]
37
[ "coco2017/train2017/000000000338.jpg", "coco2017/depth/000000000338_depth.png" ]
[ { "from": "human", "value": "<image>Predict relative depth visualized as a grayscale image with subtle, continuous shading transitions. The output should show gradual value changes throughout, not discrete color regions like segmentation masks." }, { "from": "gpt", "value": "<image>" } ]
38
[ "coco2017/train2017/000000000357.jpg", "coco2017/depth/000000000357_depth.png" ]
[ { "from": "human", "value": "<image>Visualize relative depth as a continuous grayscale image. The output should show gradual value transitions without discrete boundaries, distinguishing it from segmentation blocky color regions." }, { "from": "gpt", "value": "<image>" } ]
39
[ "coco2017/train2017/000000000368.jpg", "coco2017/depth/000000000368_depth.png" ]
[ { "from": "human", "value": "<image>Output a grayscale image with continuous tone variations representing depth. Unlike segmentation discrete color blocks, this should show smooth brightness gradients across the entire image." }, { "from": "gpt", "value": "<image>" } ]
40
[ "coco2017/train2017/000000000370.jpg", "coco2017/depth/000000000370_depth.png" ]
[ { "from": "human", "value": "<image>Estimate relative depth per pixel. Output a grayscale image where brighter pixels represent closer distances. The result should show continuous tone variations without discrete color jumps, distinguishing it from other task outputs." }, { "from": "gpt", "value...
41
[ "coco2017/train2017/000000000384.jpg", "coco2017/depth/000000000384_depth.png" ]
[ { "from": "human", "value": "<image>Visualize relative depth as a continuous grayscale image. The output should show gradual value transitions without discrete boundaries, distinguishing it from segmentation blocky color regions." }, { "from": "gpt", "value": "<image>" } ]
42
[ "coco2017/train2017/000000000387.jpg", "coco2017/depth/000000000387_depth.png" ]
[ { "from": "human", "value": "<image>Predict per-pixel relative depth, visualized as a continuous grayscale image where brightness indicates proximity (brighter for closer, darker for farther). The output should be a smooth gray image with varying shades of gray, representing a continuous depth field." }, ...
43
[ "coco2017/train2017/000000000389.jpg", "coco2017/depth/000000000389_depth.png" ]
[ { "from": "human", "value": "<image>Predict relative depth visualized as a grayscale image with subtle, continuous shading transitions. The output should show gradual value changes throughout, not discrete color regions like segmentation masks." }, { "from": "gpt", "value": "<image>" } ]
44
[ "coco2017/train2017/000000000394.jpg", "coco2017/depth/000000000394_depth.png" ]
[ { "from": "human", "value": "<image>Output a grayscale image with continuous tone variations representing depth. Unlike segmentation discrete color blocks, this should show smooth brightness gradients across the entire image." }, { "from": "gpt", "value": "<image>" } ]
45
[ "coco2017/train2017/000000000395.jpg", "coco2017/depth/000000000395_depth.png" ]
[ { "from": "human", "value": "<image>Predict per-pixel relative depth, visualized as a continuous grayscale image where brightness indicates proximity (brighter for closer, darker for farther). The output should be a smooth gray image with varying shades of gray, representing a continuous depth field." }, ...
46
[ "coco2017/train2017/000000000397.jpg", "coco2017/depth/000000000397_depth.png" ]
[ { "from": "human", "value": "<image>Generate a continuous depth map as a grayscale image with smooth variations in gray levels (0-255). Closer regions appear brighter, farther regions appear darker, creating a visually continuous depth representation. Output is a image" }, { "from": "gpt", "valu...
47
[ "coco2017/train2017/000000000400.jpg", "coco2017/depth/000000000400_depth.png" ]
[ { "from": "human", "value": "<image>Generate a continuous depth map as a grayscale image with smooth variations in gray levels (0-255). Closer regions appear brighter, farther regions appear darker, creating a visually continuous depth representation. Output is a image" }, { "from": "gpt", "valu...
48
[ "coco2017/train2017/000000000415.jpg", "coco2017/depth/000000000415_depth.png" ]
[ { "from": "human", "value": "<image>Output a grayscale image with continuous tone variations representing depth. Unlike segmentation discrete color blocks, this should show smooth brightness gradients across the entire image." }, { "from": "gpt", "value": "<image>" } ]
49
[ "coco2017/train2017/000000000428.jpg", "coco2017/depth/000000000428_depth.png" ]
[ { "from": "human", "value": "<image>Predict relative depth visualized as a grayscale image with subtle, continuous shading transitions. The output should show gradual value changes throughout, not discrete color regions like segmentation masks." }, { "from": "gpt", "value": "<image>" } ]
50
[ "coco2017/train2017/000000000431.jpg", "coco2017/depth/000000000431_depth.png" ]
[ { "from": "human", "value": "<image>Predict relative depth visualized as a grayscale image with subtle, continuous shading transitions. The output should show gradual value changes throughout, not discrete color regions like segmentation masks." }, { "from": "gpt", "value": "<image>" } ]
51
[ "coco2017/train2017/000000000436.jpg", "coco2017/depth/000000000436_depth.png" ]
[ { "from": "human", "value": "<image>Generate a continuous depth map as a grayscale image with smooth variations in gray levels (0-255). Closer regions appear brighter, farther regions appear darker, creating a visually continuous depth representation. Output is a image" }, { "from": "gpt", "valu...
52
[ "coco2017/train2017/000000000438.jpg", "coco2017/depth/000000000438_depth.png" ]
[ { "from": "human", "value": "<image>Predict per-pixel relative depth, visualized as a continuous grayscale image where brightness indicates proximity (brighter for closer, darker for farther). The output should be a smooth gray image with varying shades of gray, representing a continuous depth field." }, ...
53
[ "coco2017/train2017/000000000443.jpg", "coco2017/depth/000000000443_depth.png" ]
[ { "from": "human", "value": "<image>Visualize relative depth as a continuous grayscale image. The output should show gradual value transitions without discrete boundaries, distinguishing it from segmentation blocky color regions." }, { "from": "gpt", "value": "<image>" } ]
54
[ "coco2017/train2017/000000000446.jpg", "coco2017/depth/000000000446_depth.png" ]
[ { "from": "human", "value": "<image>Predict per-pixel relative depth, visualized as a continuous grayscale image where brightness indicates proximity (brighter for closer, darker for farther). The output should be a smooth gray image with varying shades of gray, representing a continuous depth field." }, ...
55
[ "coco2017/train2017/000000000450.jpg", "coco2017/depth/000000000450_depth.png" ]
[ { "from": "human", "value": "<image>Estimate relative depth per pixel. Output a grayscale image where brighter pixels represent closer distances. The result should show continuous tone variations without discrete color jumps, distinguishing it from other task outputs." }, { "from": "gpt", "value...
56
[ "coco2017/train2017/000000000459.jpg", "coco2017/depth/000000000459_depth.png" ]
[ { "from": "human", "value": "<image>Estimate relative depth for each pixel in the image, with closer objects appearing brighter and distant objects appearing darker. Output is a grayscale image with pixel values ranging from 0-255." }, { "from": "gpt", "value": "<image>" } ]
57
[ "coco2017/train2017/000000000474.jpg", "coco2017/depth/000000000474_depth.png" ]
[ { "from": "human", "value": "<image>Estimate relative depth for each pixel in the image, with closer objects appearing brighter and distant objects appearing darker. Output is a grayscale image with pixel values ranging from 0-255." }, { "from": "gpt", "value": "<image>" } ]
58
[ "coco2017/train2017/000000000486.jpg", "coco2017/depth/000000000486_depth.png" ]
[ { "from": "human", "value": "<image>Estimate relative depth for each pixel in the image, with closer objects appearing brighter and distant objects appearing darker. Output is a grayscale image with pixel values ranging from 0-255." }, { "from": "gpt", "value": "<image>" } ]
59
[ "coco2017/train2017/000000000490.jpg", "coco2017/depth/000000000490_depth.png" ]
[ { "from": "human", "value": "<image>Generate a continuous depth map as a grayscale image with smooth variations in gray levels (0-255). Closer regions appear brighter, farther regions appear darker, creating a visually continuous depth representation. Output is a image" }, { "from": "gpt", "valu...
60
[ "coco2017/train2017/000000000502.jpg", "coco2017/depth/000000000502_depth.png" ]
[ { "from": "human", "value": "<image>Predict per-pixel relative depth, visualized as a continuous grayscale image where brightness indicates proximity (brighter for closer, darker for farther). The output should be a smooth gray image with varying shades of gray, representing a continuous depth field." }, ...
61
[ "coco2017/train2017/000000000510.jpg", "coco2017/depth/000000000510_depth.png" ]
[ { "from": "human", "value": "<image>Visualize relative depth as a continuous grayscale image. The output should show gradual value transitions without discrete boundaries, distinguishing it from segmentation blocky color regions." }, { "from": "gpt", "value": "<image>" } ]
62
[ "coco2017/train2017/000000000514.jpg", "coco2017/depth/000000000514_depth.png" ]
[ { "from": "human", "value": "<image>Visualize relative depth as a continuous grayscale image. The output should show gradual value transitions without discrete boundaries, distinguishing it from segmentation blocky color regions." }, { "from": "gpt", "value": "<image>" } ]
63
[ "coco2017/train2017/000000000536.jpg", "coco2017/depth/000000000536_depth.png" ]
[ { "from": "human", "value": "<image>Predict per-pixel relative depth, visualized as a continuous grayscale image where brightness indicates proximity (brighter for closer, darker for farther). The output should be a smooth gray image with varying shades of gray, representing a continuous depth field." }, ...
64
[ "coco2017/train2017/000000000544.jpg", "coco2017/depth/000000000544_depth.png" ]
[ { "from": "human", "value": "<image>Output a grayscale image with continuous tone variations representing depth. Unlike segmentation discrete color blocks, this should show smooth brightness gradients across the entire image." }, { "from": "gpt", "value": "<image>" } ]
65
[ "coco2017/train2017/000000000560.jpg", "coco2017/depth/000000000560_depth.png" ]
[ { "from": "human", "value": "<image>Estimate relative depth for each pixel in the image, with closer objects appearing brighter and distant objects appearing darker. Output is a grayscale image with pixel values ranging from 0-255." }, { "from": "gpt", "value": "<image>" } ]
66
[ "coco2017/train2017/000000000562.jpg", "coco2017/depth/000000000562_depth.png" ]
[ { "from": "human", "value": "<image>Visualize relative depth as a continuous grayscale image. The output should show gradual value transitions without discrete boundaries, distinguishing it from segmentation blocky color regions." }, { "from": "gpt", "value": "<image>" } ]
67
[ "coco2017/train2017/000000000564.jpg", "coco2017/depth/000000000564_depth.png" ]
[ { "from": "human", "value": "<image>Estimate relative depth for each pixel in the image, with closer objects appearing brighter and distant objects appearing darker. Output is a grayscale image with pixel values ranging from 0-255." }, { "from": "gpt", "value": "<image>" } ]
68
[ "coco2017/train2017/000000000569.jpg", "coco2017/depth/000000000569_depth.png" ]
[ { "from": "human", "value": "<image>Visualize relative depth as a continuous grayscale image. The output should show gradual value transitions without discrete boundaries, distinguishing it from segmentation blocky color regions." }, { "from": "gpt", "value": "<image>" } ]
69
[ "coco2017/train2017/000000000572.jpg", "coco2017/depth/000000000572_depth.png" ]
[ { "from": "human", "value": "<image>Predict relative depth visualized as a grayscale image with subtle, continuous shading transitions. The output should show gradual value changes throughout, not discrete color regions like segmentation masks." }, { "from": "gpt", "value": "<image>" } ]
70
[ "coco2017/train2017/000000000575.jpg", "coco2017/depth/000000000575_depth.png" ]
[ { "from": "human", "value": "<image>Output a grayscale image with continuous tone variations representing depth. Unlike segmentation discrete color blocks, this should show smooth brightness gradients across the entire image." }, { "from": "gpt", "value": "<image>" } ]
71
[ "coco2017/train2017/000000000581.jpg", "coco2017/depth/000000000581_depth.png" ]
[ { "from": "human", "value": "<image>Visualize relative depth as a continuous grayscale image. The output should show gradual value transitions without discrete boundaries, distinguishing it from segmentation blocky color regions." }, { "from": "gpt", "value": "<image>" } ]
72
[ "coco2017/train2017/000000000584.jpg", "coco2017/depth/000000000584_depth.png" ]
[ { "from": "human", "value": "<image>Estimate relative depth for each pixel in the image, with closer objects appearing brighter and distant objects appearing darker. Output is a grayscale image with pixel values ranging from 0-255." }, { "from": "gpt", "value": "<image>" } ]
73
[ "coco2017/train2017/000000000590.jpg", "coco2017/depth/000000000590_depth.png" ]
[ { "from": "human", "value": "<image>Estimate relative depth per pixel. Output a grayscale image where brighter pixels represent closer distances. The result should show continuous tone variations without discrete color jumps, distinguishing it from other task outputs." }, { "from": "gpt", "value...
74
[ "coco2017/train2017/000000000599.jpg", "coco2017/depth/000000000599_depth.png" ]
[ { "from": "human", "value": "<image>Predict relative depth visualized as a grayscale image with subtle, continuous shading transitions. The output should show gradual value changes throughout, not discrete color regions like segmentation masks." }, { "from": "gpt", "value": "<image>" } ]
75
[ "coco2017/train2017/000000000605.jpg", "coco2017/depth/000000000605_depth.png" ]
[ { "from": "human", "value": "<image>Output a grayscale image with continuous tone variations representing depth. Unlike segmentation discrete color blocks, this should show smooth brightness gradients across the entire image." }, { "from": "gpt", "value": "<image>" } ]
76
[ "coco2017/train2017/000000000612.jpg", "coco2017/depth/000000000612_depth.png" ]
[ { "from": "human", "value": "<image>Predict relative depth visualized as a grayscale image with subtle, continuous shading transitions. The output should show gradual value changes throughout, not discrete color regions like segmentation masks." }, { "from": "gpt", "value": "<image>" } ]
77
[ "coco2017/train2017/000000000620.jpg", "coco2017/depth/000000000620_depth.png" ]
[ { "from": "human", "value": "<image>Estimate relative depth for each pixel in the image, with closer objects appearing brighter and distant objects appearing darker. Output is a grayscale image with pixel values ranging from 0-255." }, { "from": "gpt", "value": "<image>" } ]
78
[ "coco2017/train2017/000000000623.jpg", "coco2017/depth/000000000623_depth.png" ]
[ { "from": "human", "value": "<image>Predict relative depth visualized as a grayscale image with subtle, continuous shading transitions. The output should show gradual value changes throughout, not discrete color regions like segmentation masks." }, { "from": "gpt", "value": "<image>" } ]
79
[ "coco2017/train2017/000000000625.jpg", "coco2017/depth/000000000625_depth.png" ]
[ { "from": "human", "value": "<image>Output a grayscale image with continuous tone variations representing depth. Unlike segmentation discrete color blocks, this should show smooth brightness gradients across the entire image." }, { "from": "gpt", "value": "<image>" } ]
80
[ "coco2017/train2017/000000000629.jpg", "coco2017/depth/000000000629_depth.png" ]
[ { "from": "human", "value": "<image>Visualize relative depth as a continuous grayscale image. The output should show gradual value transitions without discrete boundaries, distinguishing it from segmentation blocky color regions." }, { "from": "gpt", "value": "<image>" } ]
81
[ "coco2017/train2017/000000000634.jpg", "coco2017/depth/000000000634_depth.png" ]
[ { "from": "human", "value": "<image>Predict relative depth visualized as a grayscale image with subtle, continuous shading transitions. The output should show gradual value changes throughout, not discrete color regions like segmentation masks." }, { "from": "gpt", "value": "<image>" } ]
82
[ "coco2017/train2017/000000000636.jpg", "coco2017/depth/000000000636_depth.png" ]
[ { "from": "human", "value": "<image>Generate a continuous depth map as a grayscale image with smooth variations in gray levels (0-255). Closer regions appear brighter, farther regions appear darker, creating a visually continuous depth representation. Output is a image" }, { "from": "gpt", "valu...
83
[ "coco2017/train2017/000000000643.jpg", "coco2017/depth/000000000643_depth.png" ]
[ { "from": "human", "value": "<image>Visualize relative depth as a continuous grayscale image. The output should show gradual value transitions without discrete boundaries, distinguishing it from segmentation blocky color regions." }, { "from": "gpt", "value": "<image>" } ]
84
[ "coco2017/train2017/000000000650.jpg", "coco2017/depth/000000000650_depth.png" ]
[ { "from": "human", "value": "<image>Generate a continuous depth map as a grayscale image with smooth variations in gray levels (0-255). Closer regions appear brighter, farther regions appear darker, creating a visually continuous depth representation. Output is a image" }, { "from": "gpt", "valu...
85
[ "coco2017/train2017/000000000656.jpg", "coco2017/depth/000000000656_depth.png" ]
[ { "from": "human", "value": "<image>Generate a continuous depth map as a grayscale image with smooth variations in gray levels (0-255). Closer regions appear brighter, farther regions appear darker, creating a visually continuous depth representation. Output is a image" }, { "from": "gpt", "valu...
86
[ "coco2017/train2017/000000000661.jpg", "coco2017/depth/000000000661_depth.png" ]
[ { "from": "human", "value": "<image>Generate a continuous depth map as a grayscale image with smooth variations in gray levels (0-255). Closer regions appear brighter, farther regions appear darker, creating a visually continuous depth representation. Output is a image" }, { "from": "gpt", "valu...
87
[ "coco2017/train2017/000000000670.jpg", "coco2017/depth/000000000670_depth.png" ]
[ { "from": "human", "value": "<image>Estimate relative depth for each pixel in the image, with closer objects appearing brighter and distant objects appearing darker. Output is a grayscale image with pixel values ranging from 0-255." }, { "from": "gpt", "value": "<image>" } ]
88
[ "coco2017/train2017/000000000673.jpg", "coco2017/depth/000000000673_depth.png" ]
[ { "from": "human", "value": "<image>Visualize relative depth as a continuous grayscale image. The output should show gradual value transitions without discrete boundaries, distinguishing it from segmentation blocky color regions." }, { "from": "gpt", "value": "<image>" } ]
89
[ "coco2017/train2017/000000000675.jpg", "coco2017/depth/000000000675_depth.png" ]
[ { "from": "human", "value": "<image>Predict relative depth visualized as a grayscale image with subtle, continuous shading transitions. The output should show gradual value changes throughout, not discrete color regions like segmentation masks." }, { "from": "gpt", "value": "<image>" } ]
90
[ "coco2017/train2017/000000000684.jpg", "coco2017/depth/000000000684_depth.png" ]
[ { "from": "human", "value": "<image>Estimate relative depth per pixel. Output a grayscale image where brighter pixels represent closer distances. The result should show continuous tone variations without discrete color jumps, distinguishing it from other task outputs." }, { "from": "gpt", "value...
91
[ "coco2017/train2017/000000000692.jpg", "coco2017/depth/000000000692_depth.png" ]
[ { "from": "human", "value": "<image>Predict relative depth visualized as a grayscale image with subtle, continuous shading transitions. The output should show gradual value changes throughout, not discrete color regions like segmentation masks." }, { "from": "gpt", "value": "<image>" } ]
92
[ "coco2017/train2017/000000000693.jpg", "coco2017/depth/000000000693_depth.png" ]
[ { "from": "human", "value": "<image>Predict per-pixel relative depth, visualized as a continuous grayscale image where brightness indicates proximity (brighter for closer, darker for farther). The output should be a smooth gray image with varying shades of gray, representing a continuous depth field." }, ...
93
[ "coco2017/train2017/000000000699.jpg", "coco2017/depth/000000000699_depth.png" ]
[ { "from": "human", "value": "<image>Output a grayscale image with continuous tone variations representing depth. Unlike segmentation discrete color blocks, this should show smooth brightness gradients across the entire image." }, { "from": "gpt", "value": "<image>" } ]
94
[ "coco2017/train2017/000000000711.jpg", "coco2017/depth/000000000711_depth.png" ]
[ { "from": "human", "value": "<image>Predict per-pixel relative depth, visualized as a continuous grayscale image where brightness indicates proximity (brighter for closer, darker for farther). The output should be a smooth gray image with varying shades of gray, representing a continuous depth field." }, ...
95
[ "coco2017/train2017/000000000714.jpg", "coco2017/depth/000000000714_depth.png" ]
[ { "from": "human", "value": "<image>Estimate relative depth for each pixel in the image, with closer objects appearing brighter and distant objects appearing darker. Output is a grayscale image with pixel values ranging from 0-255." }, { "from": "gpt", "value": "<image>" } ]
96
[ "coco2017/train2017/000000000715.jpg", "coco2017/depth/000000000715_depth.png" ]
[ { "from": "human", "value": "<image>Predict relative depth visualized as a grayscale image with subtle, continuous shading transitions. The output should show gradual value changes throughout, not discrete color regions like segmentation masks." }, { "from": "gpt", "value": "<image>" } ]
97
[ "coco2017/train2017/000000000716.jpg", "coco2017/depth/000000000716_depth.png" ]
[ { "from": "human", "value": "<image>Estimate relative depth per pixel. Output a grayscale image where brighter pixels represent closer distances. The result should show continuous tone variations without discrete color jumps, distinguishing it from other task outputs." }, { "from": "gpt", "value...
98
[ "coco2017/train2017/000000000723.jpg", "coco2017/depth/000000000723_depth.png" ]
[ { "from": "human", "value": "<image>Generate a continuous depth map as a grayscale image with smooth variations in gray levels (0-255). Closer regions appear brighter, farther regions appear darker, creating a visually continuous depth representation. Output is a image" }, { "from": "gpt", "valu...
99
[ "coco2017/train2017/000000000730.jpg", "coco2017/depth/000000000730_depth.png" ]
[ { "from": "human", "value": "<image>Generate a continuous depth map as a grayscale image with smooth variations in gray levels (0-255). Closer regions appear brighter, farther regions appear darker, creating a visually continuous depth representation. Output is a image" }, { "from": "gpt", "valu...
End of preview. Expand in Data Studio
README.md exists but content is empty.
Downloads last month
160