index
stringlengths 44
47
| question
stringlengths 88
244
| A
stringlengths 5
161
| B
stringlengths 4
49
| C
stringclasses 1
value | D
stringclasses 1
value | answer
stringclasses 3
values | answer_cot
stringlengths 583
1.78k
| answer_name
stringlengths 8
165
| category
stringlengths 13
37
| image_source
stringclasses 3
values | image_url
stringlengths 40
126
|
|---|---|---|---|---|---|---|---|---|---|---|---|
KITTI_train_55ac744a11b74c0ea4473a168162684d | Consider the real-world 3D locations and orientations of the objects. Which side of The car is white is facing This is a car? | front | left | back | right | B | To solve this problem, we first detect the 3D locations of The car is white and This is a car. Then we compute the vector pointing from The car is white to This is a car. Now we compute the angles between the vector and the left, right, front, back directions. We first compute the left direction of The car is white, which leads to the angle between left direction and the vector. Immediately we also get the angle between the right direction and the vector. Similarly we compute the angles between the vector and front, back directions. The side of The car is white that is facing This is a car corresponds to the direction with the smallest angle. Based on the four estimated angles, we can predict the final answer. The 3D location of The car is white is (-8.5, 1.0, 29.5). The 3D location of This is a car is (-8.5, 1.0, 23.9). The vector from The car is white to This is a car is hence (-0.0, -0.0, -5.6). The left direction of The car is white is (-0.0, 0.0, -0.8). The cosine similarity between the vector pointing to This is a car and the left direction is 1.00, corresponding to an angle of 1.98 degrees. Thus the angle between the vector pointing to This is a car and the right direction is 178.02 degrees. The front direction of The car is white is (-2.0, 0.0, 0.1). The cosine similarity between the vector pointing to This is a car and the front direction is -0.03, corresponding to an angle of 91.97 degrees. Thus the angle between the vector pointing to This is a car and the back direction is 88.03 degrees. Among the four directions, the smallest angle is the left direction, with an angle of 1.98 degrees. Thus the left side of The car is white is facing the This is a car. Therefore, the final answer is B. left. | B. left. | multi_object_viewpoint_towards_object | KITTI | KITTI_object/training/image_2/000470.png |
KITTI_train_8d238a2e2214462991691f4af9147bf3 | Consider the real-world 3D locations and orientations of the objects. If I stand at The car is white's position facing where it is facing, is The car is white in front of me or behind me? | in front of | behind | null | null | B | To solve this problem, we first determine the 3D locations of The car is white and The car is white. Then we estimate the vector pointing from The car is white to The car is white, as well as the front direction of The car is white. Next we compute the cosine similarities between the vector and the front direction, which leads to the angle between the vector and the fron direction. If the angle is smaller than 90 degrees, then The car is white is in front of The car is white. Otherwise, The car is white is behind The car is white. The 3D location of The car is white is (-7.2, 1.0, 10.2). The 3D location of The car is white is (-9.0, 1.3, 23.0). The vector from The car is white to The car is white is hence (1.8, -0.3, -12.8). The front direction of The car is white is (0.3, 0.0, 2.1). The cosine similarity between the vector and the front direction is -0.96, corresponding to an angle of 164.33 degrees. The angle is smaller than 90 degrees, meaning that The car is white is behind The car is white. Therefore, the final answer is B. behind. | B. behind. | orientation_in_front_of | KITTI | KITTI_object/training/image_2/000817.png |
KITTI_train_3c4b71ecc09346ac98f7dac4e1e121c8 | Consider the real-world 3D orientations of the objects. Are A white van parked on the side of the road and A yellow car parked on the side of the road facing same or similar directions, or very different directions? | same or similar directions | very different directions | null | null | B | To solve this problem, we first detect the front directions of A white van parked on the side of the road and A yellow car parked on the side of the road. Then we compute the cosine similarities between the two front directions, and the angle between them. If the angle between the two front directions is small, then the two objects are facing same or similar directions. Otherwise, the two objects are facing very different directions. The front direction of A white van parked on the side of the road is (3.3, 0.0, 0.0). The front direction of A yellow car parked on the side of the road is (-0.0, 0.0, 1.6). The cosine similarity between the two front directions is -0.01, corresponding to an angle of 90.53. The angle is large, meaning that the two objects are facing very different directions. Therefore, the final answer is B. very different directions. | B. very different directions. | multi_object_same_direction | KITTI | KITTI_object/training/image_2/001476.png |
KITTI_train_d59ed298ed1b44d9bf2ad1427fd4c0c2 | Consider the real-world 3D location of the objects. Which object is further away from the camera? | The truck is parked on the side of the road | The car is red | null | null | A | To solve this problem, we first estimate the 3D locations of The truck is parked on the side of the road and The car is red. Then we estimate the L2 distances from the camera to the two objects. The object with a smaller L2 distance is the one that is closer to the camera. The 3D location of The truck is parked on the side of the road is (-6.3, 0.7, 31.5). The 3D location of The car is red is (6.6, 0.8, 13.4). The L2 distance from the camera to The truck is parked on the side of the road is 32.13. The L2 distance from the camera to The car is red is 14.98. The distance to The truck is parked on the side of the road is larger. Therefore, the answer is A. The truck is parked on the side of the road. | A. The truck is parked on the side of the road. | location_closer_to_camera | KITTI | KITTI_object/training/image_2/001838.png |
KITTI_train_d6eddaedc3584c30adc99f80bde6ef9b | Consider the real-world 3D locations of the objects. Which is closer to The car is red and white, The car is red or The car is red? | The car is red | The car is red | null | null | A | To solve this problem, we first detect the 3D location of The car is red and white, The car is red, and The car is red. Then we compute the L2 distances between The car is red and white and The car is red, and between The car is red and white and The car is red. The object that is closer to The car is red and white is the one with a smaller distance. The 3D location of The car is red and white is (0.2, 1.0, 14.6). The 3D location of The car is red is (15.6, 1.2, 20.5). The 3D location of The car is red is (-18.5, 2.1, 27.8). The L2 distance between The car is red and white and The car is red is 16.504520897994187. The L2 distance between The car is red and white and The car is red is 22.966095806406532. Between the two distances, the distance between The car is red and white and The car is red is smaller. Therefore, the final answer is A. The car is red. | A. The car is red. | multi_object_closer_to | KITTI | KITTI_object/training/image_2/003639.png |
KITTI_train_59079da3f86c410ebdfdc240d3a5f280 | Consider the real-world 3D orientations of the objects. What is the relationship between the orientations of The car is red and A car parked on the side of the road, parallel of perpendicular to each other? | parallel | perpendicular | null | null | B | To solve this problem, we first detect the front directions of The car is red and A car parked on the side of the road. Then we compute the cosine similarities between the two front directions, and the angle between them. If the angle between the two front directions is closer to 0 or 180, than to 90 degrees, then the two objects are parallel to each other. Otherwise, the two objects are perpendicular to each other. The front direction of The car is red is (0.0, 0.0, -1.8). The front direction of A car parked on the side of the road is (-2.1, 0.0, -0.1). The cosine similarity between the two front directions is 0.05, corresponding to an angle of 87.09. The angle is closer to 90, than to 0 or 180 degrees, meaning that the two objects are perpendicular to each other. Therefore, the final answer is B. perpendicular. | B. perpendicular. | multi_object_parallel | KITTI | KITTI_object/training/image_2/004102.png |
KITTI_train_b941944b06be4041aa2ebf04239071b3 | Consider the real-world 3D locations and orientations of the objects. If I stand at The car is black's position facing where it is facing, is A black car parked on the side of the road in front of me or behind me? | in front of | behind | null | null | A | To solve this problem, we first determine the 3D locations of A black car parked on the side of the road and The car is black. Then we estimate the vector pointing from The car is black to A black car parked on the side of the road, as well as the front direction of The car is black. Next we compute the cosine similarities between the vector and the front direction, which leads to the angle between the vector and the fron direction. If the angle is smaller than 90 degrees, then A black car parked on the side of the road is in front of The car is black. Otherwise, A black car parked on the side of the road is behind The car is black. The 3D location of A black car parked on the side of the road is (3.1, 0.9, 14.0). The 3D location of The car is black is (2.9, 0.9, 9.2). The vector from The car is black to A black car parked on the side of the road is hence (0.1, 0.0, 4.8). The front direction of The car is black is (0.0, 0.0, 2.0). The cosine similarity between the vector and the front direction is 1.00, corresponding to an angle of 0.56 degrees. The angle is smaller than 90 degrees, meaning that A black car parked on the side of the road is in front of The car is black. Therefore, the final answer is A. in front of. | A. in front of. | orientation_in_front_of | KITTI | KITTI_object/training/image_2/005959.png |
KITTI_train_aa03cf4d244247528e8289f2c00b7088 | Consider the real-world 3D locations of the objects. Are the The car is silver and the A white car parked on the side of the road next to each other or far away from each other? | next to each other | far away from each other | null | null | B | To solve this problem, we first estimate the 3D locations of The car is silver and A white car parked on the side of the road. Then we can compute the L2 distance between the two objects. Next we estimate the rough sizes of the two objects. If the distance between the two objects is smaller or roughly the same as the object sizes, then the two objects are next to each other. Otherwise, the two objects are far away from each other. The 3D location of The car is silver is (-5.0, 1.3, 15.3). The 3D location of A white car parked on the side of the road is (-5.0, 1.3, 20.4). The L2 distance between the two objects is 5.04. The size of the The car is silver is roughly 3.94. The size of the A white car parked on the side of the road is roughly 3.90. The distance between the two objects is much bigger than the object sizes, meaning that they are far away from each other. Therefore, the answer is B. far away from each other. | B. far away from each other. | location_next_to | KITTI | KITTI_object/training/image_2/006025.png |
KITTI_train_fc911dabb0af4ccd9dbe0fcfad0b04d5 | Consider the real-world 3D locations of the objects. Are the A car is parked on the side of the road and the A person walking on the sidewalk next to each other or far away from each other? | next to each other | far away from each other | null | null | B | To solve this problem, we first estimate the 3D locations of A car is parked on the side of the road and A person walking on the sidewalk. Then we can compute the L2 distance between the two objects. Next we estimate the rough sizes of the two objects. If the distance between the two objects is smaller or roughly the same as the object sizes, then the two objects are next to each other. Otherwise, the two objects are far away from each other. The 3D location of A car is parked on the side of the road is (3.6, 0.7, 5.5). The 3D location of A person walking on the sidewalk is (7.1, 0.5, 9.7). The L2 distance between the two objects is 5.45. The size of the A car is parked on the side of the road is roughly 4.12. The size of the A person walking on the sidewalk is roughly 1.72. The distance between the two objects is much bigger than the object sizes, meaning that they are far away from each other. Therefore, the answer is B. far away from each other. | B. far away from each other. | location_next_to | KITTI | KITTI_object/training/image_2/007203.png |
KITTI_train_4668b4d66e06458c9c4331f040712dc1 | Consider the real-world 3D location of the objects. Which object is further away from the camera? | A car on the street | The van is white | null | null | B | To solve this problem, we first estimate the 3D locations of A car on the street and The van is white. Then we estimate the L2 distances from the camera to the two objects. The object with a smaller L2 distance is the one that is closer to the camera. The 3D location of A car on the street is (-5.7, 1.2, 9.6). The 3D location of The van is white is (-3.5, 1.0, 21.9). The L2 distance from the camera to A car on the street is 11.22. The L2 distance from the camera to The van is white is 22.17. The distance to The van is white is larger. Therefore, the answer is B. The van is white. | B. The van is white. | location_closer_to_camera | KITTI | KITTI_object/training/image_2/007301.png |
SUNRGBD_train_a48a92a4c2f44a9687bd2d4dc47f62d0 | Consider the real-world 3D orientations of the objects. What is the relationship between the orientations of A computer monitor with a black frame and The chair is black and red, parallel of perpendicular to each other? | parallel | perpendicular | null | null | A | To solve this problem, we first detect the front directions of A computer monitor with a black frame and The chair is black and red. Then we compute the cosine similarities between the two front directions, and the angle between them. If the angle between the two front directions is closer to 0 or 180, than to 90 degrees, then the two objects are parallel to each other. Otherwise, the two objects are perpendicular to each other. The front direction of A computer monitor with a black frame is (-0.0, 0.0, -0.0). The front direction of The chair is black and red is (0.3, -0.1, 0.3). The cosine similarity between the two front directions is -1.00, corresponding to an angle of 177.54. The angle is closer to 0 or 180, than to 90 degrees, meaning that the two objects are parallel to each other. Therefore, the final answer is A. parallel. | A. parallel. | multi_object_parallel | SUNRGBD | SUNRGBD/kv1/b3dodata/img_0098//image/img_0098.jpg |
SUNRGBD_train_3c5a2e47d7a546f097f60f7b2c429ac8 | Consider the real-world 3D locations of the objects. Are the The table is brown and the The chair is brown next to each other or far away from each other? | next to each other | far away from each other | null | null | A | To solve this problem, we first estimate the 3D locations of The table is brown and The chair is brown. Then we can compute the L2 distance between the two objects. Next we estimate the rough sizes of the two objects. If the distance between the two objects is smaller or roughly the same as the object sizes, then the two objects are next to each other. Otherwise, the two objects are far away from each other. The 3D location of The table is brown is (-0.1, 0.2, 2.4). The 3D location of The chair is brown is (-0.5, 0.3, 2.3). The L2 distance between the two objects is 0.44. The size of the The table is brown is roughly 1.06. The size of the The chair is brown is roughly 1.12. The distance between the two objects is smaller or roughly the same as the object sizes, meaning that they are next to each other. Therefore, the answer is A. next to each other. | A. next to each other. | location_next_to | SUNRGBD | SUNRGBD/kv2/align_kv2/2014-12-18_16-25-58_260595134347//image/0000031.jpg |
SUNRGBD_train_1fc96e5c469d44f8a7d2f506ffc7f62d | Consider the real-world 3D locations of the objects. Which is closer to The bed is purple, A bed with a purple comforter or A purple bed with a black frame? | A bed with a purple comforter | A purple bed with a black frame | null | null | B | To solve this problem, we first detect the 3D location of The bed is purple, A bed with a purple comforter, and A purple bed with a black frame. Then we compute the L2 distances between The bed is purple and A bed with a purple comforter, and between The bed is purple and A purple bed with a black frame. The object that is closer to The bed is purple is the one with a smaller distance. The 3D location of The bed is purple is (-0.1, 0.3, 2.2). The 3D location of A bed with a purple comforter is (1.8, -1.5, 5.5). The 3D location of A purple bed with a black frame is (2.0, -0.6, 4.5). The L2 distance between The bed is purple and A bed with a purple comforter is 4.189616005039021. The L2 distance between The bed is purple and A purple bed with a black frame is 3.189270742707693. Between the two distances, the distance between The bed is purple and A purple bed with a black frame is smaller. Therefore, the final answer is B. A purple bed with a black frame. | B. A purple bed with a black frame. | multi_object_closer_to | SUNRGBD | SUNRGBD/kv2/kinect2data/002378_2014-06-28_20-14-24_260595134347_rgbf000030-resize/image/0000030.jpg |
SUNRGBD_train_533aba0ae7c24fae885c291cb64a1925 | Consider the real-world 3D locations and orientations of the objects. If I stand at The bed is white and has a wooden frame's position facing where it is facing, is A yellow pillow with a blue arrow pointing to it on the left or right of me? | on the left | on the right | null | null | A | To solve this problem, we first determine the 3D locations of A yellow pillow with a blue arrow pointing to it and The bed is white and has a wooden frame. Then we estimate the vector pointing from The bed is white and has a wooden frame to A yellow pillow with a blue arrow pointing to it, as well as the left direction of The bed is white and has a wooden frame. Next we compute the cosine similarities between the vector and the left direction, which leads to the angle between the vector and the fron direction. If the angle is smaller than 90 degrees, then A yellow pillow with a blue arrow pointing to it is on the left of The bed is white and has a wooden frame. Otherwise, A yellow pillow with a blue arrow pointing to it is behind The bed is white and has a wooden frame. The 3D location of A yellow pillow with a blue arrow pointing to it is (0.5, -0.5, 2.6). The 3D location of The bed is white and has a wooden frame is (-1.9, -1.1, 4.5). The vector from The bed is white and has a wooden frame to A yellow pillow with a blue arrow pointing to it is hence (2.4, 0.6, -1.9). The left direction of The bed is white and has a wooden frame is (0.6, 0.2, -0.7). The cosine similarity between the vector and the left direction is 0.98, corresponding to an angle of 12.18 degrees. The angle is smaller than 90 degrees, meaning that A yellow pillow with a blue arrow pointing to it is on the left of The bed is white and has a wooden frame. Therefore, the final answer is A. on the left. | A. on the left. | orientation_on_the_left | SUNRGBD | SUNRGBD/kv2/kinect2data/002403_2014-06-28_20-22-02_260595134347_rgbf000045-resize/image/0000045.jpg |
SUNRGBD_train_ddefd05764e1455ba884b34ad0077377 | Consider the real-world 3D locations and orientations of the objects. If I stand at The bed is a twin bed with a blue and white striped comforter's position facing where it is facing, is The pillow is blue and white in front of me or behind me? | in front of | behind | null | null | B | To solve this problem, we first determine the 3D locations of The pillow is blue and white and The bed is a twin bed with a blue and white striped comforter. Then we estimate the vector pointing from The bed is a twin bed with a blue and white striped comforter to The pillow is blue and white, as well as the front direction of The bed is a twin bed with a blue and white striped comforter. Next we compute the cosine similarities between the vector and the front direction, which leads to the angle between the vector and the fron direction. If the angle is smaller than 90 degrees, then The pillow is blue and white is in front of The bed is a twin bed with a blue and white striped comforter. Otherwise, The pillow is blue and white is behind The bed is a twin bed with a blue and white striped comforter. The 3D location of The pillow is blue and white is (-1.3, -0.3, 3.0). The 3D location of The bed is a twin bed with a blue and white striped comforter is (-0.6, 0.2, 2.7). The vector from The bed is a twin bed with a blue and white striped comforter to The pillow is blue and white is hence (-0.6, -0.5, 0.3). The front direction of The bed is a twin bed with a blue and white striped comforter is (1.0, 0.1, -0.2). The cosine similarity between the vector and the front direction is -0.85, corresponding to an angle of 147.95 degrees. The angle is smaller than 90 degrees, meaning that The pillow is blue and white is behind The bed is a twin bed with a blue and white striped comforter. Therefore, the final answer is B. behind. | B. behind. | orientation_in_front_of | SUNRGBD | SUNRGBD/kv2/kinect2data/003050_2014-06-15_13-50-46_094959634447_rgbf000150-resize/image/0000150.jpg |
SUNRGBD_train_5257f859409642049548a5280ba674b2 | Consider the real-world 3D orientations of the objects. What is the relationship between the orientations of The chair is tan in color and The chair is made of wood and has a brown color, parallel of perpendicular to each other? | parallel | perpendicular | null | null | A | To solve this problem, we first detect the front directions of The chair is tan in color and The chair is made of wood and has a brown color. Then we compute the cosine similarities between the two front directions, and the angle between them. If the angle between the two front directions is closer to 0 or 180, than to 90 degrees, then the two objects are parallel to each other. Otherwise, the two objects are perpendicular to each other. The front direction of The chair is tan in color is (0.1, -0.1, 0.2). The front direction of The chair is made of wood and has a brown color is (-0.1, 0.1, -0.2). The cosine similarity between the two front directions is -0.99, corresponding to an angle of 173.95. The angle is closer to 0 or 180, than to 90 degrees, meaning that the two objects are parallel to each other. Therefore, the final answer is A. parallel. | A. parallel. | multi_object_parallel | SUNRGBD | SUNRGBD/realsense/sa/2014_10_24-21_24_08-1311000073//image/0000060.jpg |
SUNRGBD_train_23f4fedb6eba431bb961cb40b1c3359b | Consider the real-world 3D orientations of the objects. Are This is a chair and The chair is white and gray facing same or similar directions, or very different directions? | same or similar directions | very different directions | null | null | A | To solve this problem, we first detect the front directions of This is a chair and The chair is white and gray. Then we compute the cosine similarities between the two front directions, and the angle between them. If the angle between the two front directions is small, then the two objects are facing same or similar directions. Otherwise, the two objects are facing very different directions. The front direction of This is a chair is (-0.1, 0.1, -0.3). The front direction of The chair is white and gray is (-0.1, 0.1, -0.3). The cosine similarity between the two front directions is 1.00, corresponding to an angle of 1.31. The angle is small, meaning that the two objects are facing same or similar directions. Therefore, the final answer is A. same or similar directions. | A. same or similar directions. | multi_object_same_direction | SUNRGBD | SUNRGBD/xtion/sun3ddata/harvard_c7/hv_c7_1/0000645-000021584304//image/0000645-000021584304.jpg |
SUNRGBD_train_b71d92a70f67413e9ce1d7d632868f2a | Consider the real-world 3D location of the objects. Which object is closer to the camera? | A white trash can | A white trash can | null | null | A | To solve this problem, we first estimate the 3D locations of A white trash can and A white trash can. Then we estimate the L2 distances from the camera to the two objects. The object with a smaller L2 distance is the one that is closer to the camera. The 3D location of A white trash can is (0.5, 0.2, 1.1). The 3D location of A white trash can is (-0.8, -0.2, 1.9). The L2 distance from the camera to A white trash can is 1.23. The L2 distance from the camera to A white trash can is 2.05. The distance to A white trash can is smaller. Therefore, the answer is A. A white trash can. | A. A white trash can. | location_closer_to_camera | SUNRGBD | SUNRGBD/xtion/sun3ddata/home_ac/home_ac_scan2_2012_aug_22/0005274-001008776978//image/0005274-001008776978.jpg |
SUNRGBD_train_6671620de01b4aa9a2ee00e804e0f05d | Consider the real-world 3D locations of the objects. Which object has a higher location? | A shelf is visible in the image | The bed is white and has a pink and black blanket | null | null | A | To solve this problem, we first detect the 3D locations of the two objects. To determine which object is higher, we first compute the vector pointing from A shelf is visible in the image to The bed is white and has a pink and black blanket. Then we determine the up direction of A shelf is visible in the image and estimate the cosine similarity between the vector and the up direction, which leads to the angle between the two directions. If the angle between the two directions is larger than 90 degrees, this means that A shelf is visible in the image is at a higher location. Otherwise, The bed is white and has a pink and black blanket is at a higher location. The 3D location of A shelf is visible in the image is (0.4, -1.1, 3.6). The 3D location of The bed is white and has a pink and black blanket is (1.1, -0.2, 3.0). The vector from A shelf is visible in the image to The bed is white and has a pink and black blanket is hence (0.7, 0.9, -0.6). The up direction of A shelf is visible in the image is (-0.0, -0.6, -0.3). The cosine similarity between the vector and the up direction is -0.44, corresponding to an angle of 116 degrees. The angle between the vector and the up direction is larger than 90 degrees, meaning that the location of A shelf is visible in the image is higher. Therefore, the answer is A. The bed is white and has a pink and black blanket. | A. The bed is white and has a pink and black blanket. | height_higher | SUNRGBD | SUNRGBD/xtion/sun3ddata/hotel_beijing/beijing_hotel_1/0011971-000401220036//image/0011971-000401220036.jpg |
SUNRGBD_train_5502034bdab849f8a62bb8722d99941c | Consider the real-world 3D location of the objects. Which object is closer to the camera? | The night stand is brown in color and has a wooden texture | The nightstand is brown and made of wood | null | null | B | To solve this problem, we first estimate the 3D locations of The night stand is brown in color and has a wooden texture and The nightstand is brown and made of wood. Then we estimate the L2 distances from the camera to the two objects. The object with a smaller L2 distance is the one that is closer to the camera. The 3D location of The night stand is brown in color and has a wooden texture is (-1.2, -0.6, 3.9). The 3D location of The nightstand is brown and made of wood is (1.2, -0.3, 2.7). The L2 distance from the camera to The night stand is brown in color and has a wooden texture is 4.11. The L2 distance from the camera to The nightstand is brown and made of wood is 3.02. The distance to The nightstand is brown and made of wood is smaller. Therefore, the answer is B. The nightstand is brown and made of wood. | B. The nightstand is brown and made of wood. | location_closer_to_camera | SUNRGBD | SUNRGBD/xtion/sun3ddata/hotel_pittsburg/hotel_pittsburg_scan1_2012_dec_12/0016176-000542292291//image/0016176-000542292291.jpg |
nuScenes_train_a90f75826bf44e88a9b63329429a92a6 | Consider the real-world 3D locations and orientations of the objects. Which object is A person walking on the sidewalk facing towards, A person walking on the sidewalk or the A person is walking across the street? | A person walking on the sidewalk | A person is walking across the street | null | null | A | To solve this problem, we first detect the 3D location of A person walking on the sidewalk, A person walking on the sidewalk, and A person is walking across the street. Then we compute the cosine similarities between the front direction of A person walking on the sidewalk and the vectors from A person walking on the sidewalk to the other two objects. We can estimate the angles from the cosine similarities, and the object with a smaller angle is the object that A person walking on the sidewalk is facing towards. The 3D location of A person walking on the sidewalk is (5.1, 0.4, 9.9). The 3D location of A person walking on the sidewalk is (5.2, 0.3, 11.4). The 3D location of A person is walking across the street is (5.8, 0.2, 9.8). The front direction of A person walking on the sidewalk is (0.1, -0.0, 0.3). First we consider if A person walking on the sidewalk is facing towards the A person walking on the sidewalk. The vector from A person walking on the sidewalk to A person walking on the sidewalk is (0.1, -0.1, 1.5). The cosine similarity between the front direction and the vector is 0.99, corresponding to an angle of 8.99 degrees. First we consider if A person walking on the sidewalk is facing towards the A person is walking across the street. The vector from A person walking on the sidewalk to A person is walking across the street is (0.7, -0.2, -0.1). The cosine similarity between the front direction and the vector is 0.05, corresponding to an angle of 86.95 degrees. We find that the angle between the front direction and A person walking on the sidewalk is smaller. Therefore, the final answer is A. A person walking on the sidewalk. | A. A person walking on the sidewalk. | multi_object_facing | nuScenes | nuScenes/samples/CAM_FRONT/n008-2018-08-21-11-53-44-0400__CAM_FRONT__1534867308912404.jpg |
nuScenes_train_e0b8ff848cd04caa97b30bebc73c8deb | Consider the real-world 3D locations and orientations of the objects. If I stand at The car is dark in color's position facing where it is facing, is The pedestrian is walking on the sidewalk in front of me or behind me? | in front of | behind | null | null | A | To solve this problem, we first determine the 3D locations of The pedestrian is walking on the sidewalk and The car is dark in color. Then we estimate the vector pointing from The car is dark in color to The pedestrian is walking on the sidewalk, as well as the front direction of The car is dark in color. Next we compute the cosine similarities between the vector and the front direction, which leads to the angle between the vector and the fron direction. If the angle is smaller than 90 degrees, then The pedestrian is walking on the sidewalk is in front of The car is dark in color. Otherwise, The pedestrian is walking on the sidewalk is behind The car is dark in color. The 3D location of The pedestrian is walking on the sidewalk is (-4.9, 0.7, 8.3). The 3D location of The car is dark in color is (7.2, 0.6, 17.7). The vector from The car is dark in color to The pedestrian is walking on the sidewalk is hence (-12.1, 0.1, -9.5). The front direction of The car is dark in color is (-2.0, 0.0, -1.3). The cosine similarity between the vector and the front direction is 1.00, corresponding to an angle of 4.33 degrees. The angle is smaller than 90 degrees, meaning that The pedestrian is walking on the sidewalk is in front of The car is dark in color. Therefore, the final answer is A. in front of. | A. in front of. | orientation_in_front_of | nuScenes | nuScenes/samples/CAM_FRONT/n008-2018-08-28-13-40-50-0400__CAM_FRONT__1535478555012404.jpg |
nuScenes_train_358ccb06464147acb72916e629d8766f | Consider the real-world 3D locations and orientations of the objects. If I stand at The truck is white's position facing where it is facing, is A person walking on the sidewalk in front of me or behind me? | in front of | behind | null | null | A | To solve this problem, we first determine the 3D locations of A person walking on the sidewalk and The truck is white. Then we estimate the vector pointing from The truck is white to A person walking on the sidewalk, as well as the front direction of The truck is white. Next we compute the cosine similarities between the vector and the front direction, which leads to the angle between the vector and the fron direction. If the angle is smaller than 90 degrees, then A person walking on the sidewalk is in front of The truck is white. Otherwise, A person walking on the sidewalk is behind The truck is white. The 3D location of A person walking on the sidewalk is (4.4, 0.6, 7.5). The 3D location of The truck is white is (-14.8, -0.2, 44.2). The vector from The truck is white to A person walking on the sidewalk is hence (19.3, 0.8, -36.7). The front direction of The truck is white is (0.1, -0.0, -3.3). The cosine similarity between the vector and the front direction is 0.90, corresponding to an angle of 25.30 degrees. The angle is smaller than 90 degrees, meaning that A person walking on the sidewalk is in front of The truck is white. Therefore, the final answer is A. in front of. | A. in front of. | orientation_in_front_of | nuScenes | nuScenes/samples/CAM_FRONT/n008-2018-08-31-11-19-57-0400__CAM_FRONT__1535729154412404.jpg |
nuScenes_train_92da3eb350104148b901a8223c0a20ba | Consider the real-world 3D location of the objects. Which object is closer to the camera? | A black car is parked in a parking lot | A person walking on the sidewalk | null | null | B | To solve this problem, we first estimate the 3D locations of A black car is parked in a parking lot and A person walking on the sidewalk. Then we estimate the L2 distances from the camera to the two objects. The object with a smaller L2 distance is the one that is closer to the camera. The 3D location of A black car is parked in a parking lot is (-9.2, 0.4, 29.1). The 3D location of A person walking on the sidewalk is (-5.5, 0.5, 15.7). The L2 distance from the camera to A black car is parked in a parking lot is 30.57. The L2 distance from the camera to A person walking on the sidewalk is 16.63. The distance to A person walking on the sidewalk is smaller. Therefore, the answer is B. A person walking on the sidewalk. | B. A person walking on the sidewalk. | location_closer_to_camera | nuScenes | nuScenes/samples/CAM_FRONT/n015-2018-08-01-16-41-59+0800__CAM_FRONT__1533112993412460.jpg |
nuScenes_train_d53469266e484b3bb2c2ee3d8e80f5a3 | Consider the real-world 3D location of the objects. Which object is further away from the camera? | The car is silver in color and has a sleek, modern design. It is driving down a street with other vehicles, including a truck, and is approaching a traffic light | The barrier is black in color | null | null | A | To solve this problem, we first estimate the 3D locations of The car is silver in color and has a sleek, modern design. It is driving down a street with other vehicles, including a truck, and is approaching a traffic light and The barrier is black in color. Then we estimate the L2 distances from the camera to the two objects. The object with a smaller L2 distance is the one that is closer to the camera. The 3D location of The car is silver in color and has a sleek, modern design. It is driving down a street with other vehicles, including a truck, and is approaching a traffic light is (-0.7, 0.5, 17.9). The 3D location of The barrier is black in color is (4.1, 0.9, 11.2). The L2 distance from the camera to The car is silver in color and has a sleek, modern design. It is driving down a street with other vehicles, including a truck, and is approaching a traffic light is 17.97. The L2 distance from the camera to The barrier is black in color is 11.94. The distance to The car is silver in color and has a sleek, modern design. It is driving down a street with other vehicles, including a truck, and is approaching a traffic light is larger. Therefore, the answer is A. The car is silver in color and has a sleek, modern design. It is driving down a street with other vehicles, including a truck, and is approaching a traffic light. | A. The car is silver in color and has a sleek, modern design. It is driving down a street with other vehicles, including a truck, and is approaching a traffic light. | location_closer_to_camera | nuScenes | nuScenes/samples/CAM_FRONT/n015-2018-09-25-11-10-38+0800__CAM_FRONT__1537845411662468.jpg |
nuScenes_train_44bad40f93314fa1994b049fc79945d3 | Consider the real-world 3D orientations of the objects. What is the relationship between the orientations of A car is parked on the side of the road and The pedestrian is wearing a black jacket, parallel of perpendicular to each other? | parallel | perpendicular | null | null | A | To solve this problem, we first detect the front directions of A car is parked on the side of the road and The pedestrian is wearing a black jacket. Then we compute the cosine similarities between the two front directions, and the angle between them. If the angle between the two front directions is closer to 0 or 180, than to 90 degrees, then the two objects are parallel to each other. Otherwise, the two objects are perpendicular to each other. The front direction of A car is parked on the side of the road is (-2.2, 0.0, 0.0). The front direction of The pedestrian is wearing a black jacket is (-2.0, 0.0, 0.1). The cosine similarity between the two front directions is 1.00, corresponding to an angle of 2.40. The angle is closer to 0 or 180, than to 90 degrees, meaning that the two objects are parallel to each other. Therefore, the final answer is A. parallel. | A. parallel. | multi_object_parallel | nuScenes | nuScenes/samples/CAM_FRONT/n015-2018-10-08-15-52-24+0800__CAM_FRONT__1538985446662460.jpg |
nuScenes_train_11f8aad97224463c9704b36234065e2d | Consider the real-world 3D orientations of the objects. What is the relationship between the orientations of A person is walking on the sidewalk and The car is white, parallel of perpendicular to each other? | parallel | perpendicular | null | null | A | To solve this problem, we first detect the front directions of A person is walking on the sidewalk and The car is white. Then we compute the cosine similarities between the two front directions, and the angle between them. If the angle between the two front directions is closer to 0 or 180, than to 90 degrees, then the two objects are parallel to each other. Otherwise, the two objects are perpendicular to each other. The front direction of A person is walking on the sidewalk is (-0.8, 0.1, 2.2). The front direction of The car is white is (-0.8, 0.1, 1.9). The cosine similarity between the two front directions is 1.00, corresponding to an angle of 2.87. The angle is closer to 0 or 180, than to 90 degrees, meaning that the two objects are parallel to each other. Therefore, the final answer is A. parallel. | A. parallel. | multi_object_parallel | nuScenes | nuScenes/samples/CAM_FRONT/n015-2018-10-08-16-03-24+0800__CAM_FRONT__1538985947612460.jpg |
nuScenes_train_d95f4a0fd11c472cbb0e3753ca49e2e4 | Consider the real-world 3D locations and orientations of the objects. Which side of The car is white is facing The bus is orange? | front | left | back | right | A | To solve this problem, we first detect the 3D locations of The car is white and The bus is orange. Then we compute the vector pointing from The car is white to The bus is orange. Now we compute the angles between the vector and the left, right, front, back directions. We first compute the left direction of The car is white, which leads to the angle between left direction and the vector. Immediately we also get the angle between the right direction and the vector. Similarly we compute the angles between the vector and front, back directions. The side of The car is white that is facing The bus is orange corresponds to the direction with the smallest angle. Based on the four estimated angles, we can predict the final answer. The 3D location of The car is white is (0.5, 0.7, 13.7). The 3D location of The bus is orange is (7.1, 0.1, 35.2). The vector from The car is white to The bus is orange is hence (6.6, -0.6, 21.5). The left direction of The car is white is (-0.9, -0.0, -0.0). The cosine similarity between the vector pointing to The bus is orange and the left direction is -0.30, corresponding to an angle of 107.25 degrees. Thus the angle between the vector pointing to The bus is orange and the right direction is 72.75 degrees. The front direction of The car is white is (-0.0, 0.0, 2.2). The cosine similarity between the vector pointing to The bus is orange and the front direction is 0.95, corresponding to an angle of 17.35 degrees. Thus the angle between the vector pointing to The bus is orange and the back direction is 162.65 degrees. Among the four directions, the smallest angle is the front direction, with an angle of 17.35 degrees. Thus the front side of The car is white is facing the The bus is orange. Therefore, the final answer is A. front. | A. front. | multi_object_viewpoint_towards_object | nuScenes | nuScenes/samples/CAM_FRONT/n015-2018-10-08-16-03-24+0800__CAM_FRONT__1538985963612460.jpg |
nuScenes_train_fd36c86c67ac447b8800cd9d78483e17 | Consider the real-world 3D locations of the objects. Which is closer to A person is standing on the sidewalk, A car on the road or The car is white? | A car on the road | The car is white | null | null | B | To solve this problem, we first detect the 3D location of A person is standing on the sidewalk, A car on the road, and The car is white. Then we compute the L2 distances between A person is standing on the sidewalk and A car on the road, and between A person is standing on the sidewalk and The car is white. The object that is closer to A person is standing on the sidewalk is the one with a smaller distance. The 3D location of A person is standing on the sidewalk is (-9.4, 0.4, 21.1). The 3D location of A car on the road is (4.3, 0.7, 8.9). The 3D location of The car is white is (4.5, 0.8, 16.6). The L2 distance between A person is standing on the sidewalk and A car on the road is 18.33694343468256. The L2 distance between A person is standing on the sidewalk and The car is white is 14.639591288577982. Between the two distances, the distance between A person is standing on the sidewalk and The car is white is smaller. Therefore, the final answer is B. The car is white. | B. The car is white. | multi_object_closer_to | nuScenes | nuScenes/samples/CAM_FRONT/n015-2018-11-21-19-11-29+0800__CAM_FRONT__1542798957412460.jpg |
nuScenes_train_c5ba059c46ec438c8d106ce00508a3cf | Consider the real-world 3D locations and orientations of the objects. Which side of The person is walking on the sidewalk is facing the camera? | front | left | back | right | C | To solve this problem, we first estimate the 3D location of The person is walking on the sidewalk. Then we obtain the vector pointing from the object to the camera. Now we compute the angles between the vector and the left, right, front, back directions. We first compute the left direction of The person is walking on the sidewalk, which leads to the angle between left direction and the vector. Immediately we also get the angle between the right direction and the vector. Similarly we compute the angles between the vector and front, back directions. The side of The person is walking on the sidewalk that is facing the camera corresponds to the direction with the smallest angle. Based on the four estimated angles, we can predict the final answer. The location of The person is walking on the sidewalk is (10.8, 0.1, 20.5). The vector from The person is walking on the sidewalk to camera is hence (-10.8, -0.1, -20.5). The left direction of The person is walking on the sidewalk is (-0.9, -0.0, 0.1). The cosine similarity between the vector pointing to camera and the left direction is 0.36, corresponding to an angle of 68.83 degrees. Thus the angle between the vector pointing to camera and the right direction is 111.17 degrees. The front direction of The person is walking on the sidewalk is (0.3, -0.0, 2.5). The cosine similarity between the vector pointing to camera and the front direction is -0.93, corresponding to an angle of 158.83 degrees. Thus the angle between the vector pointing to camera and the back direction is 21.17 degrees. Among the four directions, the smallest angle is the back direction, with an angle of 21.17 degrees. Thus the back side of The person is walking on the sidewalk is facing the camera. Therefore, the final answer is C. back. | C. back. | orientation_viewpoint | nuScenes | nuScenes/samples/CAM_FRONT/n015-2018-11-21-19-38-26+0800__CAM_FRONT__1542801295912460.jpg |
README.md exists but content is empty.
- Downloads last month
- 110