image_path
stringlengths 12
96
| image_label
stringclasses 527
values | question
stringlengths 50
148
| question_type
stringclasses 5
values | answer_choices
listlengths 5
5
| answer_types
listlengths 5
5
| answer_parts
listlengths 5
5
| segmentations
dict | object_question
stringclasses 1
value | object_answer_choices
listlengths 5
5
⌀ | object_answer_types
listlengths 5
5
⌀ | object_answer_classes
listlengths 5
5
⌀ |
|---|---|---|---|---|---|---|---|---|---|---|---|
images/0efcede787a8fd36be19111e9867f7c7757fcef2cfcbcec27bb1c68eb016b839.jpg
|
agricultural airplane
|
What visible parts does the agricultural airplane in the image have?
|
identification_with_label
| ["The agricultural airplane in the image has the following visible parts: fixed landing gear, propul(...TRUNCATED)
|
[
"correct",
"incorrect",
"incorrect",
"incorrect",
"incorrect"
] | [["fixed landing gear","propulsion component","spraying rig","wings"],["fixed landing gear","propuls(...TRUNCATED)
| {"abrasive belt":null,"abrasive surface":null,"absorbent layer":null,"address panel":null,"adhesive (...TRUNCATED)
| null | null | null | null |
images/0a6d54eeb3a6cc41b1ba9187fb6fd7e51b4103b0f4cf13decc26f5b668631e01.jpg
|
agricultural airplane
|
What visible parts does the agricultural airplane in the image have?
|
identification_with_label
| ["The agricultural airplane in the image has the following visible parts: fixed landing gear, propul(...TRUNCATED)
|
[
"correct",
"incorrect",
"incorrect",
"incorrect",
"incorrect"
] | [["fixed landing gear","propulsion component","spraying rig","wings"],["propulsion component","spray(...TRUNCATED)
| {"abrasive belt":null,"abrasive surface":null,"absorbent layer":null,"address panel":null,"adhesive (...TRUNCATED)
| null | null | null | null |
images/douglas_a4_1.jpg
|
attack aircraft
|
What visible parts does the attack aircraft in the image have?
|
identification_with_label
| ["The attack aircraft in the image has the following visible parts: bomb, bomb racks, propulsion com(...TRUNCATED)
|
[
"correct",
"incorrect",
"incorrect",
"incorrect",
"incorrect"
] | [["bomb","bomb racks","propulsion component","wings"],["blades","bomb","propulsion component","wings(...TRUNCATED)
| {"abrasive belt":null,"abrasive surface":null,"absorbent layer":null,"address panel":null,"adhesive (...TRUNCATED)
| null | null | null | null |
images/emb_314_1.jpeg
|
attack aircraft
|
What visible parts does the attack aircraft in the image have?
|
identification_with_label
| ["The attack aircraft in the image has the following visible parts: bomb, bomb racks, propulsion com(...TRUNCATED)
|
[
"correct",
"incorrect",
"incorrect",
"incorrect",
"incorrect"
] | [["bomb","bomb racks","propulsion component","wings"],["bomb","propulsion component","wings"],["bomb(...TRUNCATED)
| {"abrasive belt":null,"abrasive surface":null,"absorbent layer":null,"address panel":null,"adhesive (...TRUNCATED)
| null | null | null | null |
images/0e03098a2925013ea695183b1c2c7340aab7ba7dacd1c1f595aa1287d3961b28.jpg
|
bomber aircraft
|
What visible parts does the bomber aircraft in the image have?
|
identification_with_label
| ["The bomber aircraft in the image has the following visible parts: propulsion component, and wings.(...TRUNCATED)
|
[
"correct",
"incorrect",
"incorrect",
"incorrect",
"incorrect"
] | [["propulsion component","wings"],["fire retardant","load stabilizer","propulsion component"],["pres(...TRUNCATED)
| {"abrasive belt":null,"abrasive surface":null,"absorbent layer":null,"address panel":null,"adhesive (...TRUNCATED)
| null | null | null | null |
images/0f97433f03a699e2c9a3de6db657b0e16679d116459cc8d687c838ff54a812f5.jpg
|
bomber aircraft
|
What visible parts does the bomber aircraft in the image have?
|
identification_with_label
| ["The bomber aircraft in the image has the following visible parts: propulsion component, and wings.(...TRUNCATED)
|
[
"correct",
"incorrect",
"incorrect",
"incorrect",
"incorrect"
] | [["propulsion component","wings"],["wings"],["fletching","propulsion component","wings"],["abrasive (...TRUNCATED)
| {"abrasive belt":null,"abrasive surface":null,"absorbent layer":null,"address panel":null,"adhesive (...TRUNCATED)
| null | null | null | null |
images/040a1d713adffd8d74a7c843e35035e6b2c6fdff51bb7828c862bd95daff9a3a.jpg
|
business jet
|
What visible parts does the business jet in the image have?
|
identification_with_label
| ["The business jet in the image has the following visible parts: propulsion component, row of window(...TRUNCATED)
|
[
"correct",
"incorrect",
"incorrect",
"incorrect",
"incorrect"
] | [["propulsion component","row of windows","wings"],["cutoff point","propulsion component","row of wi(...TRUNCATED)
| {"abrasive belt":null,"abrasive surface":null,"absorbent layer":null,"address panel":null,"adhesive (...TRUNCATED)
| null | null | null | null |
images/0c5d100c06dc6a38008caf0b6eb244deb93a342c80b9927f7e033705079daa03.jpg
|
business jet
|
What visible parts does the business jet in the image have?
|
identification_with_label
| ["The business jet in the image has the following visible parts: propulsion component, row of window(...TRUNCATED)
|
[
"correct",
"incorrect",
"incorrect",
"incorrect",
"incorrect"
] | [["propulsion component","row of windows","wings"],["keypad","propulsion component","row of windows"(...TRUNCATED)
| {"abrasive belt":null,"abrasive surface":null,"absorbent layer":null,"address panel":null,"adhesive (...TRUNCATED)
| null | null | null | null |
images/10e129b79392b0e7a5b56dddbc2021d4b2dab3951f68b3663dc62c49cf2bbc9c.jpg
|
fighter jet
|
What visible parts does the fighter jet in the image have?
|
identification_with_label
| ["The fighter jet in the image has the following visible parts: afterburner, propulsion component, t(...TRUNCATED)
|
[
"correct",
"incorrect",
"incorrect",
"incorrect",
"incorrect"
] | [["afterburner","propulsion component","tail fin","wings"],["afterburner","oar","propulsion componen(...TRUNCATED)
| {"abrasive belt":null,"abrasive surface":null,"absorbent layer":null,"address panel":null,"adhesive (...TRUNCATED)
| null | null | null | null |
images/25ee25359ba6e616a82dd2990167e1d87be658d15d794161b9bb5a3e830b2af3.jpg
|
fighter jet
|
What visible parts does the fighter jet in the image have?
|
identification_with_label
| ["The fighter jet in the image has the following visible parts: afterburner, propulsion component, t(...TRUNCATED)
|
[
"correct",
"incorrect",
"incorrect",
"incorrect",
"incorrect"
] | [["afterburner","propulsion component","tail fin","wings"],["afterburner","dust bag","inlet","propul(...TRUNCATED)
| {"abrasive belt":null,"abrasive surface":null,"absorbent layer":null,"address panel":null,"adhesive (...TRUNCATED)
| null | null | null | null |
End of preview. Expand
in Data Studio
Repository for the Partonomy-Core benchmark introduced in PARTONOMY: Large Multimodal Models with Part-Level Visual Understanding.
Code can be found at https://github.com/anselblume/partonomy.
To download the images and evaluation metadata, install the HuggingFace CLI and run
hf download partonomy/partonomy-core --repo-type dataset --local-dir .
Unpack the images directory with tar --zst -xf images.tar.zst.
The images directory contains all images used in the Partonomy-Core evaluation. Evaluation questions, answer choices, and corresponding parts for each answer choice are contained in test-qa_pairs.json. Each question-answer JSON object contains the following fields:
image_path(str): Relative path to the image corresponding to the question.image_label(str): Label for the object in the image.question(str): The LMM prompt used for evaluation of parts.question_type(str): The type of question. These include identification_with_label (identification), positive_with_label (intersection), negative_with_label (difference), whole_to_part (Whole-to-Part), and part_to_whole (Part-to-Whole).answer_choices(list[str]): The multiple choice answer choices to evaluate the model with, using each answer choice individually as the model's response to the question prompt.answer_types(list[str]): The correctness of each answer choice in theanswer_choicesfield.answer_parts(list[list[str]]): List containing each of the part names sampled for each answer choice.segmentations(dict): Segmentation masks for each part, where keys are the part names. Segmentation mask objects for each part can be decoded into binary masks usingpycocotools.mask.decode.object_question(str): Analogue ofquestionLLM prompt, used for evaluation of the object label. Only present for Whole-to-Part and Part-to-Whole questions.object_answer_choices(list[str]): Answer choices for theobject_questionprompt. Only present for Whole-to-Part and Part-to-Whole questions.object_answer_types(list[str]): The correctness of each answer choice in theobject_answer_choicesfield. Only present for Whole-to-Part and Part-to-Whole questions.object_answer_classes(list[str]): The object class used for each answer choice inobject_answer_choices.
- Downloads last month
- 439