initial commit
This commit is contained in:
commit
7be61f8c6d
137 changed files with 33491 additions and 0 deletions
|
|
@ -0,0 +1,89 @@
|
|||
# coding: utf-8
|
||||
|
||||
import sys
|
||||
dataDir = '../../VQA'
|
||||
sys.path.insert(0, '%s/PythonHelperTools/vqaTools' %(dataDir))
|
||||
from vqa import VQA
|
||||
from vqaEvaluation.vqaEval import VQAEval
|
||||
import matplotlib.pyplot as plt
|
||||
import skimage.io as io
|
||||
import json
|
||||
import random
|
||||
import os
|
||||
|
||||
# set up file names and paths
|
||||
versionType ='v2_' # this should be '' when using VQA v2.0 dataset
|
||||
taskType ='OpenEnded' # 'OpenEnded' only for v2.0. 'OpenEnded' or 'MultipleChoice' for v1.0
|
||||
dataType ='mscoco' # 'mscoco' only for v1.0. 'mscoco' for real and 'abstract_v002' for abstract for v1.0.
|
||||
dataSubType ='train2014'
|
||||
annFile ='%s/Annotations/%s%s_%s_annotations.json'%(dataDir, versionType, dataType, dataSubType)
|
||||
quesFile ='%s/Questions/%s%s_%s_%s_questions.json'%(dataDir, versionType, taskType, dataType, dataSubType)
|
||||
imgDir ='%s/Images/%s/%s/' %(dataDir, dataType, dataSubType)
|
||||
resultType ='fake'
|
||||
fileTypes = ['results', 'accuracy', 'evalQA', 'evalQuesType', 'evalAnsType']
|
||||
|
||||
# An example result json file has been provided in './Results' folder.
|
||||
|
||||
[resFile, accuracyFile, evalQAFile, evalQuesTypeFile, evalAnsTypeFile] = ['%s/Results/%s%s_%s_%s_%s_%s.json'%(dataDir, versionType, taskType, dataType, dataSubType, \
|
||||
resultType, fileType) for fileType in fileTypes]
|
||||
|
||||
# create vqa object and vqaRes object
|
||||
vqa = VQA(annFile, quesFile)
|
||||
vqaRes = vqa.loadRes(resFile, quesFile)
|
||||
|
||||
# create vqaEval object by taking vqa and vqaRes
|
||||
vqaEval = VQAEval(vqa, vqaRes, n=2) #n is precision of accuracy (number of places after decimal), default is 2
|
||||
|
||||
# evaluate results
|
||||
"""
|
||||
If you have a list of question ids on which you would like to evaluate your results, pass it as a list to below function
|
||||
By default it uses all the question ids in annotation file
|
||||
"""
|
||||
vqaEval.evaluate()
|
||||
|
||||
# print accuracies
|
||||
print "\n"
|
||||
print "Overall Accuracy is: %.02f\n" %(vqaEval.accuracy['overall'])
|
||||
print "Per Question Type Accuracy is the following:"
|
||||
for quesType in vqaEval.accuracy['perQuestionType']:
|
||||
print "%s : %.02f" %(quesType, vqaEval.accuracy['perQuestionType'][quesType])
|
||||
print "\n"
|
||||
print "Per Answer Type Accuracy is the following:"
|
||||
for ansType in vqaEval.accuracy['perAnswerType']:
|
||||
print "%s : %.02f" %(ansType, vqaEval.accuracy['perAnswerType'][ansType])
|
||||
print "\n"
|
||||
# demo how to use evalQA to retrieve low score result
|
||||
evals = [quesId for quesId in vqaEval.evalQA if vqaEval.evalQA[quesId]<35] #35 is per question percentage accuracy
|
||||
if len(evals) > 0:
|
||||
print 'ground truth answers'
|
||||
randomEval = random.choice(evals)
|
||||
randomAnn = vqa.loadQA(randomEval)
|
||||
vqa.showQA(randomAnn)
|
||||
|
||||
print '\n'
|
||||
print 'generated answer (accuracy %.02f)'%(vqaEval.evalQA[randomEval])
|
||||
ann = vqaRes.loadQA(randomEval)[0]
|
||||
print "Answer: %s\n" %(ann['answer'])
|
||||
|
||||
imgId = randomAnn[0]['image_id']
|
||||
imgFilename = 'COCO_' + dataSubType + '_'+ str(imgId).zfill(12) + '.jpg'
|
||||
if os.path.isfile(imgDir + imgFilename):
|
||||
I = io.imread(imgDir + imgFilename)
|
||||
plt.imshow(I)
|
||||
plt.axis('off')
|
||||
plt.show()
|
||||
|
||||
# plot accuracy for various question types
|
||||
plt.bar(range(len(vqaEval.accuracy['perQuestionType'])), vqaEval.accuracy['perQuestionType'].values(), align='center')
|
||||
plt.xticks(range(len(vqaEval.accuracy['perQuestionType'])), vqaEval.accuracy['perQuestionType'].keys(), rotation='0',fontsize=10)
|
||||
plt.title('Per Question Type Accuracy', fontsize=10)
|
||||
plt.xlabel('Question Types', fontsize=10)
|
||||
plt.ylabel('Accuracy', fontsize=10)
|
||||
plt.show()
|
||||
|
||||
# save evaluation results to ./Results folder
|
||||
json.dump(vqaEval.accuracy, open(accuracyFile, 'w'))
|
||||
json.dump(vqaEval.evalQA, open(evalQAFile, 'w'))
|
||||
json.dump(vqaEval.evalQuesType, open(evalQuesTypeFile, 'w'))
|
||||
json.dump(vqaEval.evalAnsType, open(evalAnsTypeFile, 'w'))
|
||||
|
||||
|
|
@ -0,0 +1 @@
|
|||
author='aagrawal'
|
||||
|
|
@ -0,0 +1,192 @@
|
|||
# coding=utf-8
|
||||
|
||||
__author__='aagrawal'
|
||||
|
||||
import re
|
||||
# This code is based on the code written by Tsung-Yi Lin for MSCOCO Python API available at the following link:
|
||||
# (https://github.com/tylin/coco-caption/blob/master/pycocoevalcap/eval.py).
|
||||
import sys
|
||||
|
||||
|
||||
class VQAEval:
|
||||
def __init__(self, vqa, vqaRes, n=2):
|
||||
self.n = n
|
||||
self.accuracy = {}
|
||||
self.evalQA = {}
|
||||
self.evalQuesType = {}
|
||||
self.evalAnsType = {}
|
||||
self.vqa = vqa
|
||||
self.vqaRes = vqaRes
|
||||
self.params = {'question_id': vqa.getQuesIds()}
|
||||
self.contractions = {"aint": "ain't", "arent": "aren't", "cant": "can't", "couldve": "could've", "couldnt": "couldn't", \
|
||||
"couldn'tve": "couldn't've", "couldnt've": "couldn't've", "didnt": "didn't", "doesnt": "doesn't", "dont": "don't", "hadnt": "hadn't", \
|
||||
"hadnt've": "hadn't've", "hadn'tve": "hadn't've", "hasnt": "hasn't", "havent": "haven't", "hed": "he'd", "hed've": "he'd've", \
|
||||
"he'dve": "he'd've", "hes": "he's", "howd": "how'd", "howll": "how'll", "hows": "how's", "Id've": "I'd've", "I'dve": "I'd've", \
|
||||
"Im": "I'm", "Ive": "I've", "isnt": "isn't", "itd": "it'd", "itd've": "it'd've", "it'dve": "it'd've", "itll": "it'll", "let's": "let's", \
|
||||
"maam": "ma'am", "mightnt": "mightn't", "mightnt've": "mightn't've", "mightn'tve": "mightn't've", "mightve": "might've", \
|
||||
"mustnt": "mustn't", "mustve": "must've", "neednt": "needn't", "notve": "not've", "oclock": "o'clock", "oughtnt": "oughtn't", \
|
||||
"ow's'at": "'ow's'at", "'ows'at": "'ow's'at", "'ow'sat": "'ow's'at", "shant": "shan't", "shed've": "she'd've", "she'dve": "she'd've", \
|
||||
"she's": "she's", "shouldve": "should've", "shouldnt": "shouldn't", "shouldnt've": "shouldn't've", "shouldn'tve": "shouldn't've", \
|
||||
"somebody'd": "somebodyd", "somebodyd've": "somebody'd've", "somebody'dve": "somebody'd've", "somebodyll": "somebody'll", \
|
||||
"somebodys": "somebody's", "someoned": "someone'd", "someoned've": "someone'd've", "someone'dve": "someone'd've", \
|
||||
"someonell": "someone'll", "someones": "someone's", "somethingd": "something'd", "somethingd've": "something'd've", \
|
||||
"something'dve": "something'd've", "somethingll": "something'll", "thats": "that's", "thered": "there'd", "thered've": "there'd've", \
|
||||
"there'dve": "there'd've", "therere": "there're", "theres": "there's", "theyd": "they'd", "theyd've": "they'd've", \
|
||||
"they'dve": "they'd've", "theyll": "they'll", "theyre": "they're", "theyve": "they've", "twas": "'twas", "wasnt": "wasn't", \
|
||||
"wed've": "we'd've", "we'dve": "we'd've", "weve": "we've", "werent": "weren't", "whatll": "what'll", "whatre": "what're", \
|
||||
"whats": "what's", "whatve": "what've", "whens": "when's", "whered": "where'd", "wheres": "where's", "whereve": "where've", \
|
||||
"whod": "who'd", "whod've": "who'd've", "who'dve": "who'd've", "wholl": "who'll", "whos": "who's", "whove": "who've", "whyll": "why'll", \
|
||||
"whyre": "why're", "whys": "why's", "wont": "won't", "wouldve": "would've", "wouldnt": "wouldn't", "wouldnt've": "wouldn't've", \
|
||||
"wouldn'tve": "wouldn't've", "yall": "y'all", "yall'll": "y'all'll", "y'allll": "y'all'll", "yall'd've": "y'all'd've", \
|
||||
"y'alld've": "y'all'd've", "y'all'dve": "y'all'd've", "youd": "you'd", "youd've": "you'd've", "you'dve": "you'd've", \
|
||||
"youll": "you'll", "youre": "you're", "youve": "you've"}
|
||||
self.manualMap = { 'none': '0',
|
||||
'zero': '0',
|
||||
'one': '1',
|
||||
'two': '2',
|
||||
'three': '3',
|
||||
'four': '4',
|
||||
'five': '5',
|
||||
'six': '6',
|
||||
'seven': '7',
|
||||
'eight': '8',
|
||||
'nine': '9',
|
||||
'ten': '10'
|
||||
}
|
||||
self.articles = ['a',
|
||||
'an',
|
||||
'the'
|
||||
]
|
||||
|
||||
|
||||
self.periodStrip = re.compile("(?!<=\d)(\.)(?!\d)")
|
||||
self.commaStrip = re.compile("(\d)(\,)(\d)")
|
||||
self.punct = [';', r"/", '[', ']', '"', '{', '}',
|
||||
'(', ')', '=', '+', '\\', '_', '-',
|
||||
'>', '<', '@', '`', ',', '?', '!']
|
||||
|
||||
|
||||
def evaluate(self, quesIds=None):
|
||||
if quesIds == None:
|
||||
quesIds = [quesId for quesId in self.params['question_id']]
|
||||
gts = {}
|
||||
res = {}
|
||||
for quesId in quesIds:
|
||||
gts[quesId] = self.vqa.qa[quesId]
|
||||
res[quesId] = self.vqaRes.qa[quesId]
|
||||
|
||||
# =================================================
|
||||
# Compute accuracy
|
||||
# =================================================
|
||||
accQA = []
|
||||
accQuesType = {}
|
||||
accAnsType = {}
|
||||
# print "computing accuracy"
|
||||
step = 0
|
||||
for quesId in quesIds:
|
||||
for ansDic in gts[quesId]['answers']:
|
||||
ansDic['answer'] = ansDic['answer'].replace('\n', ' ')
|
||||
ansDic['answer'] = ansDic['answer'].replace('\t', ' ')
|
||||
ansDic['answer'] = ansDic['answer'].strip()
|
||||
resAns = res[quesId]['answer']
|
||||
resAns = resAns.replace('\n', ' ')
|
||||
resAns = resAns.replace('\t', ' ')
|
||||
resAns = resAns.strip()
|
||||
gtAcc = []
|
||||
gtAnswers = [ans['answer'] for ans in gts[quesId]['answers']]
|
||||
|
||||
if len(set(gtAnswers)) > 1:
|
||||
for ansDic in gts[quesId]['answers']:
|
||||
ansDic['answer'] = self.processPunctuation(ansDic['answer'])
|
||||
ansDic['answer'] = self.processDigitArticle(ansDic['answer'])
|
||||
resAns = self.processPunctuation(resAns)
|
||||
resAns = self.processDigitArticle(resAns)
|
||||
|
||||
for gtAnsDatum in gts[quesId]['answers']:
|
||||
otherGTAns = [item for item in gts[quesId]['answers'] if item!=gtAnsDatum]
|
||||
matchingAns = [item for item in otherGTAns if item['answer'].lower()==resAns.lower()]
|
||||
acc = min(1, float(len(matchingAns))/3)
|
||||
gtAcc.append(acc)
|
||||
quesType = gts[quesId]['question_type']
|
||||
ansType = gts[quesId]['answer_type']
|
||||
avgGTAcc = float(sum(gtAcc))/len(gtAcc)
|
||||
accQA.append(avgGTAcc)
|
||||
if quesType not in accQuesType:
|
||||
accQuesType[quesType] = []
|
||||
accQuesType[quesType].append(avgGTAcc)
|
||||
if ansType not in accAnsType:
|
||||
accAnsType[ansType] = []
|
||||
accAnsType[ansType].append(avgGTAcc)
|
||||
self.setEvalQA(quesId, avgGTAcc)
|
||||
self.setEvalQuesType(quesId, quesType, avgGTAcc)
|
||||
self.setEvalAnsType(quesId, ansType, avgGTAcc)
|
||||
if step%100 == 0:
|
||||
self.updateProgress(step/float(len(quesIds)))
|
||||
step = step + 1
|
||||
|
||||
self.setAccuracy(accQA, accQuesType, accAnsType)
|
||||
# print "Done computing accuracy"
|
||||
|
||||
def processPunctuation(self, inText):
|
||||
outText = inText
|
||||
for p in self.punct:
|
||||
if (p + ' ' in inText or ' ' + p in inText) or (re.search(self.commaStrip, inText) != None):
|
||||
outText = outText.replace(p, '')
|
||||
else:
|
||||
outText = outText.replace(p, ' ')
|
||||
outText = self.periodStrip.sub("",
|
||||
outText,
|
||||
re.UNICODE)
|
||||
return outText
|
||||
|
||||
def processDigitArticle(self, inText):
|
||||
outText = []
|
||||
tempText = inText.lower().split()
|
||||
for word in tempText:
|
||||
word = self.manualMap.setdefault(word, word)
|
||||
if word not in self.articles:
|
||||
outText.append(word)
|
||||
else:
|
||||
pass
|
||||
for wordId, word in enumerate(outText):
|
||||
if word in self.contractions:
|
||||
outText[wordId] = self.contractions[word]
|
||||
outText = ' '.join(outText)
|
||||
return outText
|
||||
|
||||
def setAccuracy(self, accQA, accQuesType, accAnsType):
|
||||
self.accuracy['overall'] = round(100*float(sum(accQA))/len(accQA), self.n)
|
||||
self.accuracy['perQuestionType'] = {quesType: round(100*float(sum(accQuesType[quesType]))/len(accQuesType[quesType]), self.n) for quesType in accQuesType}
|
||||
self.accuracy['perAnswerType'] = {ansType: round(100*float(sum(accAnsType[ansType]))/len(accAnsType[ansType]), self.n) for ansType in accAnsType}
|
||||
|
||||
def setEvalQA(self, quesId, acc):
|
||||
self.evalQA[quesId] = round(100*acc, self.n)
|
||||
|
||||
def setEvalQuesType(self, quesId, quesType, acc):
|
||||
if quesType not in self.evalQuesType:
|
||||
self.evalQuesType[quesType] = {}
|
||||
self.evalQuesType[quesType][quesId] = round(100*acc, self.n)
|
||||
|
||||
def setEvalAnsType(self, quesId, ansType, acc):
|
||||
if ansType not in self.evalAnsType:
|
||||
self.evalAnsType[ansType] = {}
|
||||
self.evalAnsType[ansType][quesId] = round(100*acc, self.n)
|
||||
|
||||
def updateProgress(self, progress):
|
||||
barLength = 20
|
||||
status = ""
|
||||
if isinstance(progress, int):
|
||||
progress = float(progress)
|
||||
if not isinstance(progress, float):
|
||||
progress = 0
|
||||
status = "error: progress var must be float\r\n"
|
||||
if progress < 0:
|
||||
progress = 0
|
||||
status = "Halt...\r\n"
|
||||
if progress >= 1:
|
||||
progress = 1
|
||||
status = "Done...\r\n"
|
||||
block = int(round(barLength*progress))
|
||||
text = "\rFinshed Percent: [{0}] {1}% {2}".format( "#"*block + "-"*(barLength-block), int(progress*100), status)
|
||||
sys.stdout.write(text)
|
||||
sys.stdout.flush()
|
||||
73
models/common/vqa_tools/VQA/PythonHelperTools/vqaDemo.py
Normal file
73
models/common/vqa_tools/VQA/PythonHelperTools/vqaDemo.py
Normal file
|
|
@ -0,0 +1,73 @@
|
|||
# coding: utf-8
|
||||
|
||||
from vqaTools.vqa import VQA
|
||||
import random
|
||||
import skimage.io as io
|
||||
import matplotlib.pyplot as plt
|
||||
import os
|
||||
|
||||
dataDir ='../../VQA'
|
||||
versionType ='v2_' # this should be '' when using VQA v2.0 dataset
|
||||
taskType ='OpenEnded' # 'OpenEnded' only for v2.0. 'OpenEnded' or 'MultipleChoice' for v1.0
|
||||
dataType ='mscoco' # 'mscoco' only for v1.0. 'mscoco' for real and 'abstract_v002' for abstract for v1.0.
|
||||
dataSubType ='train2014'
|
||||
annFile ='%s/Annotations/%s%s_%s_annotations.json'%(dataDir, versionType, dataType, dataSubType)
|
||||
quesFile ='%s/Questions/%s%s_%s_%s_questions.json'%(dataDir, versionType, taskType, dataType, dataSubType)
|
||||
imgDir = '%s/Images/%s/%s/' %(dataDir, dataType, dataSubType)
|
||||
|
||||
# initialize VQA api for QA annotations
|
||||
vqa=VQA(annFile, quesFile)
|
||||
|
||||
# load and display QA annotations for given question types
|
||||
"""
|
||||
All possible quesTypes for abstract and mscoco has been provided in respective text files in ../QuestionTypes/ folder.
|
||||
"""
|
||||
annIds = vqa.getQuesIds(quesTypes='how many');
|
||||
anns = vqa.loadQA(annIds)
|
||||
randomAnn = random.choice(anns)
|
||||
vqa.showQA([randomAnn])
|
||||
imgId = randomAnn['image_id']
|
||||
imgFilename = 'COCO_' + dataSubType + '_'+ str(imgId).zfill(12) + '.jpg'
|
||||
if os.path.isfile(imgDir + imgFilename):
|
||||
I = io.imread(imgDir + imgFilename)
|
||||
plt.imshow(I)
|
||||
plt.axis('off')
|
||||
plt.show()
|
||||
|
||||
# load and display QA annotations for given answer types
|
||||
"""
|
||||
ansTypes can be one of the following
|
||||
yes/no
|
||||
number
|
||||
other
|
||||
"""
|
||||
annIds = vqa.getQuesIds(ansTypes='yes/no');
|
||||
anns = vqa.loadQA(annIds)
|
||||
randomAnn = random.choice(anns)
|
||||
vqa.showQA([randomAnn])
|
||||
imgId = randomAnn['image_id']
|
||||
imgFilename = 'COCO_' + dataSubType + '_'+ str(imgId).zfill(12) + '.jpg'
|
||||
if os.path.isfile(imgDir + imgFilename):
|
||||
I = io.imread(imgDir + imgFilename)
|
||||
plt.imshow(I)
|
||||
plt.axis('off')
|
||||
plt.show()
|
||||
|
||||
# load and display QA annotations for given images
|
||||
"""
|
||||
Usage: vqa.getImgIds(quesIds=[], quesTypes=[], ansTypes=[])
|
||||
Above method can be used to retrieve imageIds for given question Ids or given question types or given answer types.
|
||||
"""
|
||||
ids = vqa.getImgIds()
|
||||
annIds = vqa.getQuesIds(imgIds=random.sample(ids,5));
|
||||
anns = vqa.loadQA(annIds)
|
||||
randomAnn = random.choice(anns)
|
||||
vqa.showQA([randomAnn])
|
||||
imgId = randomAnn['image_id']
|
||||
imgFilename = 'COCO_' + dataSubType + '_'+ str(imgId).zfill(12) + '.jpg'
|
||||
if os.path.isfile(imgDir + imgFilename):
|
||||
I = io.imread(imgDir + imgFilename)
|
||||
plt.imshow(I)
|
||||
plt.axis('off')
|
||||
plt.show()
|
||||
|
||||
|
|
@ -0,0 +1 @@
|
|||
__author__ = 'aagrawal'
|
||||
179
models/common/vqa_tools/VQA/PythonHelperTools/vqaTools/vqa.py
Normal file
179
models/common/vqa_tools/VQA/PythonHelperTools/vqaTools/vqa.py
Normal file
|
|
@ -0,0 +1,179 @@
|
|||
__author__ = 'aagrawal'
|
||||
__version__ = '0.9'
|
||||
|
||||
# Interface for accessing the VQA dataset.
|
||||
|
||||
# This code is based on the code written by Tsung-Yi Lin for MSCOCO Python API available at the following link:
|
||||
# (https://github.com/pdollar/coco/blob/master/PythonAPI/pycocotools/coco.py).
|
||||
|
||||
# The following functions are defined:
|
||||
# VQA - VQA class that loads VQA annotation file and prepares data structures.
|
||||
# getQuesIds - Get question ids that satisfy given filter conditions.
|
||||
# getImgIds - Get image ids that satisfy given filter conditions.
|
||||
# loadQA - Load questions and answers with the specified question ids.
|
||||
# showQA - Display the specified questions and answers.
|
||||
# loadRes - Load result file and create result object.
|
||||
|
||||
# Help on each function can be accessed by: "help(COCO.function)"
|
||||
|
||||
import json
|
||||
import datetime
|
||||
import copy
|
||||
|
||||
|
||||
class VQA:
|
||||
def __init__(self, annotation_file=None, question_file=None):
|
||||
"""
|
||||
Constructor of VQA helper class for reading and visualizing questions and answers.
|
||||
:param annotation_file (str): location of VQA annotation file
|
||||
:return:
|
||||
"""
|
||||
# load dataset
|
||||
self.dataset = {}
|
||||
self.questions = {}
|
||||
self.qa = {}
|
||||
self.qqa = {}
|
||||
self.imgToQA = {}
|
||||
if not annotation_file == None and not question_file == None:
|
||||
# print 'loading VQA annotations and questions into memory...'
|
||||
time_t = datetime.datetime.utcnow()
|
||||
dataset = json.load(open(annotation_file, 'r'))
|
||||
questions = json.load(open(question_file, 'r'))
|
||||
# print datetime.datetime.utcnow() - time_t
|
||||
self.dataset = dataset
|
||||
self.questions = questions
|
||||
self.createIndex()
|
||||
|
||||
def createIndex(self):
|
||||
imgToQA = {ann['image_id']: [] for ann in self.dataset['annotations']}
|
||||
qa = {ann['question_id']: [] for ann in self.dataset['annotations']}
|
||||
qqa = {ann['question_id']: [] for ann in self.dataset['annotations']}
|
||||
for ann in self.dataset['annotations']:
|
||||
imgToQA[ann['image_id']] += [ann]
|
||||
qa[ann['question_id']] = ann
|
||||
for ques in self.questions['questions']:
|
||||
qqa[ques['question_id']] = ques
|
||||
# print 'index created!'
|
||||
|
||||
# create class members
|
||||
self.qa = qa
|
||||
self.qqa = qqa
|
||||
self.imgToQA = imgToQA
|
||||
|
||||
def info(self):
|
||||
"""
|
||||
Print information about the VQA annotation file.
|
||||
:return:
|
||||
"""
|
||||
|
||||
# for key, value in self.datset['info'].items():
|
||||
# print '%s: %s'%(key, value)
|
||||
|
||||
def getQuesIds(self, imgIds=[], quesTypes=[], ansTypes=[]):
|
||||
"""
|
||||
Get question ids that satisfy given filter conditions. default skips that filter
|
||||
:param imgIds (int array) : get question ids for given imgs
|
||||
quesTypes (str array) : get question ids for given question types
|
||||
ansTypes (str array) : get question ids for given answer types
|
||||
:return: ids (int array) : integer array of question ids
|
||||
"""
|
||||
imgIds = imgIds if type(imgIds) == list else [imgIds]
|
||||
quesTypes = quesTypes if type(quesTypes) == list else [quesTypes]
|
||||
ansTypes = ansTypes if type(ansTypes) == list else [ansTypes]
|
||||
|
||||
if len(imgIds) == len(quesTypes) == len(ansTypes) == 0:
|
||||
anns = self.dataset['annotations']
|
||||
else:
|
||||
if not len(imgIds) == 0:
|
||||
anns = sum([self.imgToQA[imgId] for imgId in imgIds if imgId in self.imgToQA], [])
|
||||
else:
|
||||
anns = self.dataset['annotations']
|
||||
anns = anns if len(quesTypes) == 0 else [ann for ann in anns if ann['question_type'] in quesTypes]
|
||||
anns = anns if len(ansTypes) == 0 else [ann for ann in anns if ann['answer_type'] in ansTypes]
|
||||
ids = [ann['question_id'] for ann in anns]
|
||||
return ids
|
||||
|
||||
def getImgIds(self, quesIds=[], quesTypes=[], ansTypes=[]):
|
||||
"""
|
||||
Get image ids that satisfy given filter conditions. default skips that filter
|
||||
:param quesIds (int array) : get image ids for given question ids
|
||||
quesTypes (str array) : get image ids for given question types
|
||||
ansTypes (str array) : get image ids for given answer types
|
||||
:return: ids (int array) : integer array of image ids
|
||||
"""
|
||||
quesIds = quesIds if type(quesIds) == list else [quesIds]
|
||||
quesTypes = quesTypes if type(quesTypes) == list else [quesTypes]
|
||||
ansTypes = ansTypes if type(ansTypes) == list else [ansTypes]
|
||||
|
||||
if len(quesIds) == len(quesTypes) == len(ansTypes) == 0:
|
||||
anns = self.dataset['annotations']
|
||||
else:
|
||||
if not len(quesIds) == 0:
|
||||
anns = sum([self.qa[quesId] for quesId in quesIds if quesId in self.qa], [])
|
||||
else:
|
||||
anns = self.dataset['annotations']
|
||||
anns = anns if len(quesTypes) == 0 else [ann for ann in anns if ann['question_type'] in quesTypes]
|
||||
anns = anns if len(ansTypes) == 0 else [ann for ann in anns if ann['answer_type'] in ansTypes]
|
||||
ids = [ann['image_id'] for ann in anns]
|
||||
return ids
|
||||
|
||||
def loadQA(self, ids=[]):
|
||||
"""
|
||||
Load questions and answers with the specified question ids.
|
||||
:param ids (int array) : integer ids specifying question ids
|
||||
:return: qa (object array) : loaded qa objects
|
||||
"""
|
||||
if type(ids) == list:
|
||||
return [self.qa[id] for id in ids]
|
||||
elif type(ids) == int:
|
||||
return [self.qa[ids]]
|
||||
|
||||
def showQA(self, anns):
|
||||
"""
|
||||
Display the specified annotations.
|
||||
:param anns (array of object): annotations to display
|
||||
:return: None
|
||||
"""
|
||||
if len(anns) == 0:
|
||||
return 0
|
||||
for ann in anns:
|
||||
quesId = ann['question_id']
|
||||
print("Question: %s" % (self.qqa[quesId]['question']))
|
||||
for ans in ann['answers']:
|
||||
print("Answer %d: %s" % (ans['answer_id'], ans['answer']))
|
||||
|
||||
def loadRes(self, resFile, quesFile):
|
||||
"""
|
||||
Load result file and return a result object.
|
||||
:param resFile (str) : file name of result file
|
||||
:return: res (obj) : result api object
|
||||
"""
|
||||
res = VQA()
|
||||
res.questions = json.load(open(quesFile))
|
||||
res.dataset['info'] = copy.deepcopy(self.questions['info'])
|
||||
res.dataset['task_type'] = copy.deepcopy(self.questions['task_type'])
|
||||
res.dataset['data_type'] = copy.deepcopy(self.questions['data_type'])
|
||||
res.dataset['data_subtype'] = copy.deepcopy(self.questions['data_subtype'])
|
||||
res.dataset['license'] = copy.deepcopy(self.questions['license'])
|
||||
|
||||
# print 'Loading and preparing results... '
|
||||
time_t = datetime.datetime.utcnow()
|
||||
anns = json.load(open(resFile))
|
||||
assert type(anns) == list, 'results is not an array of objects'
|
||||
annsQuesIds = [ann['question_id'] for ann in anns]
|
||||
assert set(annsQuesIds) == set(self.getQuesIds()), \
|
||||
'Results do not correspond to current VQA set. Either the results do not have predictions for all question ids in annotation file or there is atleast one question id that does not belong to the question ids in the annotation file.'
|
||||
for ann in anns:
|
||||
quesId = ann['question_id']
|
||||
if res.dataset['task_type'] == 'Multiple Choice':
|
||||
assert ann['answer'] in self.qqa[quesId][
|
||||
'multiple_choices'], 'predicted answer is not one of the multiple choices'
|
||||
qaAnn = self.qa[quesId]
|
||||
ann['image_id'] = qaAnn['image_id']
|
||||
ann['question_type'] = qaAnn['question_type']
|
||||
ann['answer_type'] = qaAnn['answer_type']
|
||||
# print 'DONE (t=%0.2fs)'%((datetime.datetime.utcnow() - time_t).total_seconds())
|
||||
|
||||
res.dataset['annotations'] = anns
|
||||
res.createIndex()
|
||||
return res
|
||||
80
models/common/vqa_tools/VQA/README.md
Normal file
80
models/common/vqa_tools/VQA/README.md
Normal file
|
|
@ -0,0 +1,80 @@
|
|||
Python API and Evaluation Code for v2.0 and v1.0 releases of the VQA dataset.
|
||||
===================
|
||||
## VQA v2.0 release ##
|
||||
This release consists of
|
||||
- Real
|
||||
- 82,783 MS COCO training images, 40,504 MS COCO validation images and 81,434 MS COCO testing images (images are obtained from [MS COCO website] (http://mscoco.org/dataset/#download))
|
||||
- 443,757 questions for training, 214,354 questions for validation and 447,793 questions for testing
|
||||
- 4,437,570 answers for training and 2,143,540 answers for validation (10 per question)
|
||||
|
||||
There is only one type of task
|
||||
- Open-ended task
|
||||
|
||||
## VQA v1.0 release ##
|
||||
This release consists of
|
||||
- Real
|
||||
- 82,783 MS COCO training images, 40,504 MS COCO validation images and 81,434 MS COCO testing images (images are obtained from [MS COCO website] (http://mscoco.org/dataset/#download))
|
||||
- 248,349 questions for training, 121,512 questions for validation and 244,302 questions for testing (3 per image)
|
||||
- 2,483,490 answers for training and 1,215,120 answers for validation (10 per question)
|
||||
- Abstract
|
||||
- 20,000 training images, 10,000 validation images and 20,000 MS COCO testing images
|
||||
- 60,000 questions for training, 30,000 questions for validation and 60,000 questions for testing (3 per image)
|
||||
- 600,000 answers for training and 300,000 answers for validation (10 per question)
|
||||
|
||||
There are two types of tasks
|
||||
- Open-ended task
|
||||
- Multiple-choice task (18 choices per question)
|
||||
|
||||
## Requirements ##
|
||||
- python 2.7
|
||||
- scikit-image (visit [this page](http://scikit-image.org/docs/dev/install.html) for installation)
|
||||
- matplotlib (visit [this page](http://matplotlib.org/users/installing.html) for installation)
|
||||
|
||||
## Files ##
|
||||
./Questions
|
||||
- For v2.0, download the question files from the [VQA download page](http://www.visualqa.org/download.html), extract them and place in this folder.
|
||||
- For v1.0, both real and abstract, question files can be found on the [VQA v1 download page](http://www.visualqa.org/vqa_v1_download.html).
|
||||
- Question files from Beta v0.9 release (123,287 MSCOCO train and val images, 369,861 questions, 3,698,610 answers) can be found below
|
||||
- [training question files](http://visualqa.org/data/mscoco/prev_rel/Beta_v0.9/Questions_Train_mscoco.zip)
|
||||
- [validation question files](http://visualqa.org/data/mscoco/prev_rel/Beta_v0.9/Questions_Val_mscoco.zip)
|
||||
- Question files from Beta v0.1 release (10k MSCOCO images, 30k questions, 300k answers) can be found [here](http://visualqa.org/data/mscoco/prev_rel/Beta_v0.1/Questions_Train_mscoco.zip).
|
||||
|
||||
./Annotations
|
||||
- For v2.0, download the annotations files from the [VQA download page](http://www.visualqa.org/download.html), extract them and place in this folder.
|
||||
- For v1.0, for both real and abstract, annotation files can be found on the [VQA v1 download page](http://www.visualqa.org/vqa_v1_download.html).
|
||||
- Annotation files from Beta v0.9 release (123,287 MSCOCO train and val images, 369,861 questions, 3,698,610 answers) can be found below
|
||||
- [training annotation files](http://visualqa.org/data/mscoco/prev_rel/Beta_v0.9/Annotations_Train_mscoco.zip)
|
||||
- [validation annotation files](http://visualqa.org/data/mscoco/prev_rel/Beta_v0.9/Annotations_Val_mscoco.zip)
|
||||
- Annotation files from Beta v0.1 release (10k MSCOCO images, 30k questions, 300k answers) can be found [here](http://visualqa.org/data/mscoco/prev_rel/Beta_v0.1/Annotations_Train_mscoco.zip).
|
||||
|
||||
./Images
|
||||
- For real, create a directory with name mscoco inside this directory. For each of train, val and test, create directories with names train2014, val2014 and test2015 respectively inside mscoco directory, download respective images from [MS COCO website](http://mscoco.org/dataset/#download) and place them in respective folders.
|
||||
- For abstract, create a directory with name abstract_v002 inside this directory. For each of train, val and test, create directories with names train2015, val2015 and test2015 respectively inside abstract_v002 directory, download respective images from [VQA download page](http://www.visualqa.org/download.html) and place them in respective folders.
|
||||
|
||||
./PythonHelperTools
|
||||
- This directory contains the Python API to read and visualize the VQA dataset
|
||||
- vqaDemo.py (demo script)
|
||||
- vqaTools (API to read and visualize data)
|
||||
|
||||
./PythonEvaluationTools
|
||||
- This directory contains the Python evaluation code
|
||||
- vqaEvalDemo.py (evaluation demo script)
|
||||
- vqaEvaluation (evaluation code)
|
||||
|
||||
./Results
|
||||
- OpenEnded_mscoco_train2014_fake_results.json (an example of a fake results file for v1.0 to run the demo)
|
||||
- Visit [VQA evaluation page] (http://visualqa.org/evaluation) for more details.
|
||||
|
||||
./QuestionTypes
|
||||
- This directory contains the following lists of question types for both real and abstract questions (question types are unchanged from v1.0 to v2.0). In a list, if there are question types of length n+k and length n with the same first n words, then the question type of length n does not include questions that belong to the question type of length n+k.
|
||||
- mscoco_question_types.txt
|
||||
- abstract_v002_question_types.txt
|
||||
|
||||
## References ##
|
||||
- [VQA: Visual Question Answering](http://visualqa.org/)
|
||||
- [Microsoft COCO](http://mscoco.org/)
|
||||
|
||||
## Developers ##
|
||||
- Aishwarya Agrawal (Virginia Tech)
|
||||
- Code for API is based on [MSCOCO API code](https://github.com/pdollar/coco).
|
||||
- The format of the code for evaluation is based on [MSCOCO evaluation code](https://github.com/tylin/coco-caption).
|
||||
Loading…
Add table
Add a link
Reference in a new issue