def create_valtokcap(data_folder=DATA_FOLDER):
import gc
gc.collect()
vcap = COCO(COCO_VALID_CAP_FILE)
valimgids, tesimgids = getValimgids(), getTesimgids()
valcap = []
for i in valimgids:
valcap += vcap.imgToAnns[i]
tescap = []
for i in tesimgids:
tescap += vcap.imgToAnns[i]
vallistedCapMap = {}
for i in valcap:
vallistedCapMap[i['id']] = [
dict([('caption', i['caption']), ('image_id', i['image_id'])])
]
valtokenizedListedCapMap = PTBTokenizer().tokenize(vallistedCapMap)
teslistedCapMap = {}
for i in tescap:
teslistedCapMap[i['id']] = [
dict([('caption', i['caption']), ('image_id', i['image_id'])])
]
testokenizedListedCapMap = PTBTokenizer().tokenize(teslistedCapMap)
valtokcap = [
] #map caption ids to a map of its tokenized caption and image id
for i, j in valtokenizedListedCapMap.iteritems():
valtokcap += [(i,
dict([('caption', j[0]),
('image_id', vallistedCapMap[i][0]['image_id'])]))
]
testokcap = []
for i, j in testokenizedListedCapMap.iteritems():
testokcap += [(i,
dict([('caption', j[0]),
('image_id', teslistedCapMap[i][0]['image_id'])]))
]
f = open(data_folder + '/preprocessed/valtokcap.json', 'w')
json.dump(valtokcap, f)
f.close()
f = open(data_folder + '/preprocessed/testokcap.json', 'w')
json.dump(testokcap, f)
f.close()
def create_valtokcap(data_folder=DATA_FOLDER):
import gc
gc.collect()
vcap = COCO(COCO_VALID_CAP_FILE)
valimgids, tesimgids = getValimgids(), getTesimgids()
valcap = []
for i in valimgids:
valcap += vcap.imgToAnns[i]
tescap = []
for i in tesimgids:
tescap += vcap.imgToAnns[i]
vallistedCapMap = {}
for i in valcap:
vallistedCapMap[i['id']] = [dict([('caption',i['caption']), ('image_id', i['image_id'])])]
valtokenizedListedCapMap = PTBTokenizer().tokenize(vallistedCapMap)
teslistedCapMap = {}
for i in tescap:
teslistedCapMap[i['id']] = [dict([('caption',i['caption']), ('image_id', i['image_id'])])]
testokenizedListedCapMap = PTBTokenizer().tokenize(teslistedCapMap)
valtokcap = [] #map caption ids to a map of its tokenized caption and image id
for i, j in valtokenizedListedCapMap.iteritems():
valtokcap += [(i, dict([('caption', j[0]), ('image_id', vallistedCapMap[i][0]['image_id'])]))]
testokcap = []
for i, j in testokenizedListedCapMap.iteritems():
testokcap += [(i, dict([('caption', j[0]), ('image_id', teslistedCapMap[i][0]['image_id'])]))]
f = open(data_folder + '/preprocessed/valtokcap.json', 'w')
json.dump(valtokcap, f)
f.close()
f = open(data_folder + '/preprocessed/testokcap.json', 'w')
json.dump(testokcap, f)
f.close()
def create_tokcap(data_folder=DATA_FOLDER):
cap = COCO(COCO_TRAIN_CAP_FILE)
listedCapMap = {}
for i in cap.loadAnns(cap.getAnnIds()):
listedCapMap[i['id']] = [dict([('caption',i['caption']), ('image_id', i['image_id'])])]
tokenizedListedCapMap = PTBTokenizer().tokenize(listedCapMap)
tokcap = [] #map caption ids to a map of its tokenized caption and image id
for i, j in tokenizedListedCapMap.iteritems():
tokcap += [(i, dict([('caption', j[0]), ('image_id', listedCapMap[i][0]['image_id'])]))]
f = open(data_folder + '/preprocessed/tokcap.json', 'w')
json.dump(tokcap, f)
f.close()
def create_tokcap(data_folder=DATA_FOLDER):
cap = COCO(COCO_TRAIN_CAP_FILE)
listedCapMap = {}
for i in cap.loadAnns(cap.getAnnIds()):
listedCapMap[i['id']] = [
dict([('caption', i['caption']), ('image_id', i['image_id'])])
]
tokenizedListedCapMap = PTBTokenizer().tokenize(listedCapMap)
tokcap = [
] #map caption ids to a map of its tokenized caption and image id
for i, j in tokenizedListedCapMap.iteritems():
tokcap += [(i,
dict([('caption', j[0]),
('image_id', listedCapMap[i][0]['image_id'])]))]
f = open(data_folder + '/preprocessed/tokcap.json', 'w')
json.dump(tokcap, f)
f.close()
