Dataset Viewer
Auto-converted to Parquet Duplicate
image
imagewidth (px)
66
3.43k
task
stringclasses
16 values
primitive
stringclasses
5 values
question
stringclasses
21 values
options
listlengths
2
10
answer
stringclasses
108 values
confidence
float32
0
1
reasoning
stringlengths
4
96
existence_check
VisPresence
这张图片中是否有水域(河流、湖泊、海洋)?
[ "yes", "no" ]
no
0.95
图片中未显示河流、湖泊或海洋
time_of_day
VisChoice
这张图片最可能是在什么时间段拍摄的?
[ "dawn", "morning", "noon", "afternoon", "dusk", "night" ]
afternoon
0.85
天空阴沉,无明显阳光,但亮度适中,符合下午特征
defect_detection
VisPresence
这张图片中的主要物体是否有可见的损坏、缺陷或异常?
[ "yes", "no" ]
no
0.95
物体无明显损坏,人物和环境正常。
indoor_outdoor
VisChoice
这张图片是室内场景还是室外场景?
[ "indoor", "outdoor", "indoor_outdoor_transition" ]
indoor
0.95
背景为人工造景,类似水族馆环境
weather_recognition
VisChoice
这张图片中的天气状况是什么?
[ "sunny", "cloudy", "rainy", "snowy", "foggy", "night", "overcast" ]
night
0.8
Image appears to be taken indoors or at night.
danger_level
VisScore
请评估这张图片中场景的危险程度。
[ "none", "low", "medium", "high", "critical" ]
low
0.85
夜晚街道,车辆停放,无明显危险迹象。
scene_classification
VisChoice
这张图片的场景类型是什么?
[ "kitchen", "bedroom", "living_room", "bathroom", "office", "restaurant", "street", "park", "highway", "indoor_mall" ]
bedroom
0.95
图片中有床和床头柜,典型卧室布局
texture_material
VisChoice
这张图片中主要物体的材质是什么?
[ "metal", "wood", "fabric", "plastic", "glass", "stone", "leather", "paper", "ceramic", "rubber" ]
fabric
0.95
红色裙子和黑色上衣为布料材质
ood_detection
VisChoice_OOD
这张图片是否看起来正常、属于常见场景?如果图片异常、模糊、不自然或不属于常规场景,请回答 abnormal。
[ "normal", "abnormal", "unsure" ]
normal
0.95
厨师持披萨,场景常见且自然
scene_classification
VisChoice
这张图片的场景类型是什么?
[ "kitchen", "bedroom", "living_room", "bathroom", "office", "restaurant", "street", "park", "highway", "indoor_mall" ]
kitchen
0.95
有炉灶、水槽和橱柜,典型厨房特征
color_recognition
VisChoice
这张图片中主要物体的颜色是什么?
[ "red", "blue", "green", "yellow", "black", "white", "brown", "gray", "orange", "pink" ]
white
0.95
裙子颜色为白色,背景为粉色墙面
object_counting_vehicle
VisCount
这张图片中可以清晰看到多少辆车(包括汽车、卡车、公交车、摩托车)?
[ "zero", "one", "two", "three", "five", "many" ]
zero
1
图片中没有车辆,是饮料冰箱。
texture_material
VisChoice
这张图片中主要物体的材质是什么?
[ "metal", "wood", "fabric", "plastic", "glass", "stone", "leather", "paper", "ceramic", "rubber" ]
fabric
0.95
床单和被子为织物,是主要物体的材质
time_of_day
VisChoice
这张图片最可能是在什么时间段拍摄的?
[ "dawn", "morning", "noon", "afternoon", "dusk", "night" ]
noon
0.95
阳光明亮,天空晴朗,为正午时分。
existence_check
VisPresence
这张图片中是否有植物或花卉?
[ "yes", "no" ]
no
1
图片中只有鱼和水,无植物或花卉
danger_level
VisScore
请评估这张图片中场景的危险程度。
[ "none", "low", "medium", "high", "critical" ]
low
0.95
飞机停稳,乘客有序上下,无明显危险迹象。
texture_material
VisChoice
这张图片中主要物体的材质是什么?
[ "metal", "wood", "fabric", "plastic", "glass", "stone", "leather", "paper", "ceramic", "rubber" ]
fabric
0.95
主要物体为白色T恤,材质为面料
time_of_day
VisChoice
这张图片最可能是在什么时间段拍摄的?
[ "dawn", "morning", "noon", "afternoon", "dusk", "night" ]
morning
0.8
食物和服装适合早晨场景
danger_level
VisScore
请评估这张图片中场景的危险程度。
[ "none", "low", "medium", "high", "critical" ]
medium
0.85
道路施工区域,车辆和行人需注意安全。
color_recognition
VisChoice
这张图片中主要物体的颜色是什么?
[ "red", "blue", "green", "yellow", "black", "white", "brown", "gray", "orange", "pink" ]
blue
0.8
衬衫为蓝白格子,蓝色为主色调。
scene_classification
VisChoice
这张图片的场景类型是什么?
[ "kitchen", "bedroom", "living_room", "bathroom", "office", "restaurant", "street", "park", "highway", "indoor_mall" ]
living_room
0.8
有椅子和桌子,像阳台或休闲区,属于客厅延伸
object_counting_person
VisCount
这张图片中可以清晰看到多少人?
[ "zero", "one", "two", "three", "four", "five", "many" ]
two
0.9
车中可见两人,其他人不清晰
ood_detection
VisChoice_OOD
这张图片是否看起来正常、属于常见场景?如果图片异常、模糊、不自然或不属于常规场景,请回答 abnormal。
[ "normal", "abnormal", "unsure" ]
normal
0.95
图片展示常见场景:两辆摩托车并排停放,画面清晰。
road_condition
VisChoice
这张图片中的道路状况如何?
[ "dry_clean", "wet", "snowy", "damaged", "under_construction", "no_road" ]
dry_clean
0.95
路面无水、雪或损坏迹象,天气晴朗。
spatial_relation
VisChoice
这张图片中主要物体位于画面的哪个位置?
[ "center", "left", "right", "top", "bottom", "foreground", "background" ]
center
0.95
小熊位于画面中央,被毯子包围
hazard_detection
VisPresence
这张图片中是否存在明显的安全隐患或危险?
[ "yes", "no" ]
yes
0.8
大象站在小凳子上,平衡风险高,易摔倒。
object_category_coarse
VisChoice
这张图片中最突出的物体属于哪个类别?
[ "person", "animal", "vehicle", "furniture", "food", "electronic", "clothing", "tool", "plant", "building" ]
animal
0.95
图片中突出显示的是羊,属于animal类别
texture_material
VisChoice
这张图片中主要物体的材质是什么?
[ "metal", "wood", "fabric", "plastic", "glass", "stone", "leather", "paper", "ceramic", "rubber" ]
fabric
0.95
裤子外观为常见面料材质
hazard_detection
VisPresence
这张图片中是否存在明显的安全隐患或危险?
[ "yes", "no" ]
no
0.95
大象在自然环境中,无明显危险迹象
object_category_coarse
VisChoice
这张图片中最突出的物体属于哪个类别?
[ "person", "animal", "vehicle", "furniture", "food", "electronic", "clothing", "tool", "plant", "building" ]
clothing
0.95
图片中突出显示的是花卉图案的吊带裙和白色上衣。
spatial_relation
VisChoice
这张图片中主要物体位于画面的哪个位置?
[ "center", "left", "right", "top", "bottom", "foreground", "background" ]
center
0.95
主要物体在画面中央,即充气船和人物。
color_recognition
VisChoice
这张图片中主要物体的颜色是什么?
[ "red", "blue", "green", "yellow", "black", "white", "brown", "gray", "orange", "pink" ]
green
0.95
路牌主要颜色为绿色
existence_check
VisPresence
这张图片中是否可以看到动物?
[ "yes", "no" ]
no
1
图片中未见动物,主要是人和办公用品。
indoor_outdoor
VisChoice
这张图片是室内场景还是室外场景?
[ "indoor", "outdoor", "indoor_outdoor_transition" ]
outdoor
1
有树木草地和天空,是室外场景
danger_level
VisScore
请评估这张图片中场景的危险程度。
[ "none", "low", "medium", "high", "critical" ]
low
0.85
道路平整,车流正常,无明显危险迹象。
time_of_day
VisChoice
这张图片最可能是在什么时间段拍摄的?
[ "dawn", "morning", "noon", "afternoon", "dusk", "night" ]
dusk
0.85
背景偏暗,有剪影效果,符合傍晚光线。
existence_check
VisPresence
这张图片中是否可以看到动物?
[ "yes", "no" ]
no
0.95
图片中未见动物,主要为超市货架和商品。
spatial_relation
VisChoice
这张图片中主要物体位于画面的哪个位置?
[ "center", "left", "right", "top", "bottom", "foreground", "background" ]
center
0.95
建筑物位于画面中央,占据主要位置
object_counting_vehicle
VisCount
这张图片中可以清晰看到多少辆车(包括汽车、卡车、公交车、摩托车)?
[ "zero", "one", "two", "three", "five", "many" ]
many
0.95
多辆车停在路边,画面中可见多辆汽车。
object_counting_vehicle
VisCount
这张图片中可以清晰看到多少辆车(包括汽车、卡车、公交车、摩托车)?
[ "zero", "one", "two", "three", "five", "many" ]
many
0.95
背景中可见多辆车辆,不只一辆
spatial_relation
VisChoice
这张图片中主要物体位于画面的哪个位置?
[ "center", "left", "right", "top", "bottom", "foreground", "background" ]
center
0.95
主要物体(汽车)位于画面中央,多人围观
indoor_outdoor
VisChoice
这张图片是室内场景还是室外场景?
[ "indoor", "outdoor", "indoor_outdoor_transition" ]
indoor
0.95
有沙发、茶几、电视,是室内客厅
spatial_relation
VisChoice
这张图片中主要物体位于画面的哪个位置?
[ "center", "left", "right", "top", "bottom", "foreground", "background" ]
center
0.95
钟表位于画面中央,背景为建筑。
object_category_coarse
VisChoice
这张图片中最突出的物体属于哪个类别?
[ "person", "animal", "vehicle", "furniture", "food", "electronic", "clothing", "tool", "plant", "building" ]
person
0.95
图像中突出显示的是穿着制服的人员。
object_counting_vehicle
VisCount
这张图片中可以清晰看到多少辆车(包括汽车、卡车、公交车、摩托车)?
[ "zero", "one", "two", "three", "five", "many" ]
many
0.95
多辆车停在路边,画面外可能还有更多
weather_recognition
VisChoice
这张图片中的天气状况是什么?
[ "sunny", "cloudy", "rainy", "snowy", "foggy", "night", "overcast" ]
sunny
0.95
阳光照射明显,天空晴朗。
spatial_relation
VisChoice
这张图片中主要物体位于画面的哪个位置?
[ "center", "left", "right", "top", "bottom", "foreground", "background" ]
center
0.95
飞机位于画面中央,占据主要位置
object_category_coarse
VisChoice
这张图片中最突出的物体属于哪个类别?
[ "person", "animal", "vehicle", "furniture", "food", "electronic", "clothing", "tool", "plant", "building" ]
person
0.95
图片中主要是人群,人是主要物体
object_counting_person
VisCount
这张图片中可以清晰看到多少人?
[ "zero", "one", "two", "three", "four", "five", "many" ]
zero
0.95
图片中未见清晰的人影。
existence_check
VisPresence
这张图片中是否有食物?
[ "yes", "no" ]
no
0.99
图片中只有海龟和水,没有食物
object_counting_vehicle
VisCount
这张图片中可以清晰看到多少辆车(包括汽车、卡车、公交车、摩托车)?
[ "zero", "one", "two", "three", "five", "many" ]
many
0.95
多辆车在街道上,包括远处和近处的车辆
scene_classification
VisChoice
这张图片的场景类型是什么?
[ "kitchen", "bedroom", "living_room", "bathroom", "office", "restaurant", "street", "park", "highway", "indoor_mall" ]
street
0.95
场景为户外土路,有大象和行人。
object_counting_vehicle
VisCount
这张图片中可以清晰看到多少辆车(包括汽车、卡车、公交车、摩托车)?
[ "zero", "one", "two", "three", "five", "many" ]
many
0.95
路边和道路上有多辆车,超过五辆
color_recognition
VisChoice
这张图片中主要物体的颜色是什么?
[ "red", "blue", "green", "yellow", "black", "white", "brown", "gray", "orange", "pink" ]
brown
0.9
房屋和地面以褐色为主
defect_detection
VisPresence
这张图片中的主要物体是否有可见的损坏、缺陷或异常?
[ "yes", "no" ]
no
0.95
肉类展示柜和环境无明显损坏迹象
object_category_coarse
VisChoice
这张图片中最突出的物体属于哪个类别?
[ "person", "animal", "vehicle", "furniture", "food", "electronic", "clothing", "tool", "plant", "building" ]
vehicle
0.95
图片中央有一辆汽车,车顶有标志,是主要物体
defect_detection
VisPresence
这张图片中的主要物体是否有可见的损坏、缺陷或异常?
[ "yes", "no" ]
no
0.95
画作和墙面无明显损坏或异常
hazard_detection
VisPresence
这张图片中是否存在明显的安全隐患或危险?
[ "yes", "no" ]
yes
0.9
设备上有'HOT'标志,存在烫伤风险
defect_detection
VisPresence
这张图片中的主要物体是否有可见的损坏、缺陷或异常?
[ "yes", "no" ]
no
0.95
设备外观无明显损坏,按钮和灯光正常显示。
danger_level
VisScore
请评估这张图片中场景的危险程度。
[ "none", "low", "medium", "high", "critical" ]
low
0.95
车辆正常行驶,未见紧急情况或违规行为。
indoor_outdoor
VisChoice
这张图片是室内场景还是室外场景?
[ "indoor", "outdoor", "indoor_outdoor_transition" ]
outdoor
0.95
有汽车和房屋,背景有树木,场景在室外
indoor_outdoor
VisChoice
这张图片是室内场景还是室外场景?
[ "indoor", "outdoor", "indoor_outdoor_transition" ]
outdoor
0.95
背景有树木和建筑物,场景在室外
scene_classification
VisChoice
这张图片的场景类型是什么?
[ "kitchen", "bedroom", "living_room", "bathroom", "office", "restaurant", "street", "park", "highway", "indoor_mall" ]
park
0.8
背景有树木,无建筑,类似公园
danger_level
VisScore
请评估这张图片中场景的危险程度。
[ "none", "low", "medium", "high", "critical" ]
low
0.95
参与者戴有护具和头盔,场景在公园,危险程度低。
texture_material
VisChoice
这张图片中主要物体的材质是什么?
[ "metal", "wood", "fabric", "plastic", "glass", "stone", "leather", "paper", "ceramic", "rubber" ]
wood
0.9
钢琴通常由木材制成,外观光滑呈木质结构。
object_category_coarse
VisChoice
这张图片中最突出的物体属于哪个类别?
[ "person", "animal", "vehicle", "furniture", "food", "electronic", "clothing", "tool", "plant", "building" ]
building
0.95
拱门是建筑结构,背景为街道场景。
danger_level
VisScore
请评估这张图片中场景的危险程度。
[ "none", "low", "medium", "high", "critical" ]
none
1
长颈鹿在动物园环境,无明显危险行为或元素
weather_recognition
VisChoice
这张图片中的天气状况是什么?
[ "sunny", "cloudy", "rainy", "snowy", "foggy", "night", "overcast" ]
overcast
0.95
天空阴沉,无阳光,符合阴天特征
spatial_relation
VisChoice
这张图片中主要物体位于画面的哪个位置?
[ "center", "left", "right", "top", "bottom", "foreground", "background" ]
center
0.95
主要物体在画面中央位置
road_condition
VisChoice
这张图片中的道路状况如何?
[ "dry_clean", "wet", "snowy", "damaged", "under_construction", "no_road" ]
dry_clean
0.95
道路干燥,无积水、积雪或损坏迹象。
weather_recognition
VisChoice
这张图片中的天气状况是什么?
[ "sunny", "cloudy", "rainy", "snowy", "foggy", "night", "overcast" ]
night
0.95
环境黑暗,场景由人工光源照亮
danger_level
VisScore
请评估这张图片中场景的危险程度。
[ "none", "low", "medium", "high", "critical" ]
medium
0.85
夜间驾驶,视线受限,路口有车辆,潜在危险。
time_of_day
VisChoice
这张图片最可能是在什么时间段拍摄的?
[ "dawn", "morning", "noon", "afternoon", "dusk", "night" ]
noon
0.95
阳光明亮,天空晴朗,影子较短,符合中午特征
texture_material
VisChoice
这张图片中主要物体的材质是什么?
[ "metal", "wood", "fabric", "plastic", "glass", "stone", "leather", "paper", "ceramic", "rubber" ]
stone
0.9
背景为石头纹理,物体在石头上。
scene_classification
VisChoice
这张图片的场景类型是什么?
[ "kitchen", "bedroom", "living_room", "bathroom", "office", "restaurant", "street", "park", "highway", "indoor_mall" ]
bedroom
0.85
窗边有床和台灯,常见于卧室
object_counting_vehicle
VisCount
这张图片中可以清晰看到多少辆车(包括汽车、卡车、公交车、摩托车)?
[ "zero", "one", "two", "three", "five", "many" ]
many
0.95
多辆车在画面中,不止三五辆
object_counting_person
VisCount
这张图片中可以清晰看到多少人?
[ "zero", "one", "two", "three", "four", "five", "many" ]
two
1
图中可见两名骑自行车的人
danger_level
VisScore
请评估这张图片中场景的危险程度。
[ "none", "low", "medium", "high", "critical" ]
medium
0.85
雨天路滑,视线不佳,需小心驾驶。
object_category_coarse
VisChoice
这张图片中最突出的物体属于哪个类别?
[ "person", "animal", "vehicle", "furniture", "food", "electronic", "clothing", "tool", "plant", "building" ]
animal
0.99
图片中是一只青蛙,属于动物类别
scene_classification
VisChoice
这张图片的场景类型是什么?
[ "kitchen", "bedroom", "living_room", "bathroom", "office", "restaurant", "street", "park", "highway", "indoor_mall" ]
bathroom
0.95
图片中有马桶和喷壶,常见于浴室。
defect_detection
VisPresence
这张图片中的主要物体是否有可见的损坏、缺陷或异常?
[ "yes", "no" ]
no
0.95
建筑外观完整,无明显损坏或缺陷
object_counting_vehicle
VisCount
这张图片中可以清晰看到多少辆车(包括汽车、卡车、公交车、摩托车)?
[ "zero", "one", "two", "three", "five", "many" ]
zero
1
图片中没有车辆。
color_recognition
VisChoice
这张图片中主要物体的颜色是什么?
[ "red", "blue", "green", "yellow", "black", "white", "brown", "gray", "orange", "pink" ]
green
0.95
青蛙主体颜色为绿色
object_counting_vehicle
VisCount
这张图片中可以清晰看到多少辆车(包括汽车、卡车、公交车、摩托车)?
[ "zero", "one", "two", "three", "five", "many" ]
many
0.95
道路上有多辆汽车,数量超过五个
time_of_day
VisChoice
这张图片最可能是在什么时间段拍摄的?
[ "dawn", "morning", "noon", "afternoon", "dusk", "night" ]
dusk
0.95
天空颜色和光线显示日落时分
object_category_coarse
VisChoice
这张图片中最突出的物体属于哪个类别?
[ "person", "animal", "vehicle", "furniture", "food", "electronic", "clothing", "tool", "plant", "building" ]
animal
0.95
图片中突出显示的是两个人骑马,马是动物。
danger_level
VisScore
请评估这张图片中场景的危险程度。
[ "none", "low", "medium", "high", "critical" ]
low
0.95
绿灯通行,车流有序,无明显危险迹象。
indoor_outdoor
VisChoice
这张图片是室内场景还是室外场景?
[ "indoor", "outdoor", "indoor_outdoor_transition" ]
indoor
0.95
有家具和室内装饰,为室内场景
object_category_coarse
VisChoice
这张图片中最突出的物体属于哪个类别?
[ "person", "animal", "vehicle", "furniture", "food", "electronic", "clothing", "tool", "plant", "building" ]
vehicle
0.95
图片中主要物体是火车,属于vehicle
time_of_day
VisChoice
这张图片最可能是在什么时间段拍摄的?
[ "dawn", "morning", "noon", "afternoon", "dusk", "night" ]
afternoon
0.9
阳光充足,影子较短,符合下午特征
object_counting_person
VisCount
这张图片中可以清晰看到多少人?
[ "zero", "one", "two", "three", "four", "five", "many" ]
many
0.95
街道上有多个行人,无法精确计数
texture_material
VisChoice
这张图片中主要物体的材质是什么?
[ "metal", "wood", "fabric", "plastic", "glass", "stone", "leather", "paper", "ceramic", "rubber" ]
wood
0.9
桌子和椅子框架为木质
color_recognition
VisChoice
这张图片中主要物体的颜色是什么?
[ "red", "blue", "green", "yellow", "black", "white", "brown", "gray", "orange", "pink" ]
blue
0.8
守门员衣服为蓝色,冰场围栏和建筑也以蓝灰色为主。
texture_material
VisChoice
这张图片中主要物体的材质是什么?
[ "metal", "wood", "fabric", "plastic", "glass", "stone", "leather", "paper", "ceramic", "rubber" ]
wood
0.95
墙壁和橱柜表面为木质纹理
object_counting_vehicle
VisCount
这张图片中可以清晰看到多少辆车(包括汽车、卡车、公交车、摩托车)?
[ "zero", "one", "two", "three", "five", "many" ]
zero
1
图片中没有车辆,只有两只鸟。
danger_level
VisScore
请评估这张图片中场景的危险程度。
[ "none", "low", "medium", "high", "critical" ]
medium
0.85
夜间驾驶,视线受限,需注意行人与车辆。
spatial_relation
VisChoice
这张图片中主要物体位于画面的哪个位置?
[ "center", "left", "right", "top", "bottom", "foreground", "background" ]
center
0.95
火车位于画面中央,占据主要位置
indoor_outdoor
VisChoice
这张图片是室内场景还是室外场景?
[ "indoor", "outdoor", "indoor_outdoor_transition" ]
outdoor
0.95
潜水场景,海龟和潜水员在水下,属于室外
texture_material
VisChoice
这张图片中主要物体的材质是什么?
[ "metal", "wood", "fabric", "plastic", "glass", "stone", "leather", "paper", "ceramic", "rubber" ]
wood
0.85
背景有大理石纹理,桌面和结构似木质
existence_check
VisPresence
这张图片中是否有人?
[ "yes", "no" ]
yes
0.9
右下角可见裤子,说明有人。
End of preview. Expand in Data Studio

ARGUS Visual Decision Dataset

A large-scale visual decision-making dataset for training and evaluating Vision-Language Models (VLMs).

Overview

  • Total samples: 16,000
  • Tasks: 16 visual decision primitives
  • Language: Chinese (zh)
  • Format: Parquet with images

Tasks

  • color_recognition: 1000
  • danger_level: 1000
  • defect_detection: 1000
  • existence_check: 1000
  • hazard_detection: 1000
  • indoor_outdoor: 1000
  • object_category_coarse: 1000
  • object_counting_person: 1000
  • object_counting_vehicle: 1000
  • ood_detection: 1000
  • road_condition: 1000
  • scene_classification: 1000
  • spatial_relation: 1000
  • texture_material: 1000
  • time_of_day: 1000
  • weather_recognition: 1000

Primitives

Primitive Description Example
VisChoice Visual multiple choice "What color is the object?" → "red"
VisScore Visual scoring/rating "Rate the damage level" → "3"
VisCount Visual counting "How many people?" → "5"
VisExist Existence verification "Is there a defect?" → "yes"

Usage

from datasets import load_dataset

ds = load_dataset("linxu/argus-visual-decision")
print(ds["train"][0])

Dataset Structure

├── data/
│   ├── train-00000-of-00001.parquet
│   └── test-00000-of-00001.parquet
└── images/
    └── <source_dataset>/<split>/<image_files>

Source Datasets

Images are sourced from public benchmarks: COCO2017, DeepFashion2, BDD100K, DOTA, Objects365, StreetHazards, ImageNet, and others.

Citation

@dataset{argus-visual-decision,
  title={ARGUS Visual Decision Dataset},
  author={Lin Xu et al.},
  year={2026}
}
Downloads last month
104