def process(options, collection, annotationName, simdir, resultfile):
rootpath = options.rootpath
if checkToSkip(resultfile, options.overwrite):
return 0
concepts = readConcepts(collection, annotationName, rootpath=rootpath)
concept_num = len(concepts)
id_images = readImageSet(collection, collection, rootpath)
image_num = len(id_images)
im2index = dict(zip(id_images, range(image_num)))
print ('%d instances, %d concepts to dump -> %s' % (image_num, concept_num, resultfile))
scores = np.zeros((image_num, concept_num)) - 1e4
for c_id,concept in enumerate(concepts):
simfile = os.path.join(simdir, '%s.txt' % concept)
ranklist = readRankingResults(simfile)
for im,score in ranklist:
idx = im2index[im]
scores[idx,c_id] = score
makedirsforfile(resultfile)
output = open(resultfile, 'wb')
pickle.dump({'concepts':concepts, 'id_images':map(int,id_images), 'scores':scores}, output, -1)
output.close()
def process(options, collection, annotationName, simdir, resultfile):
rootpath = options.rootpath
if checkToSkip(resultfile, options.overwrite):
return 0
concepts = readConcepts(collection, annotationName, rootpath=rootpath)
concept_num = len(concepts)
id_images = readImageSet(collection, collection, rootpath)
image_num = len(id_images)
im2index = dict(zip(id_images, range(image_num)))
print('%d instances, %d concepts to dump -> %s' %
(image_num, concept_num, resultfile))
scores = np.zeros((image_num, concept_num)) - 1e4
for c_id, concept in enumerate(concepts):
simfile = os.path.join(simdir, '%s.txt' % concept)
ranklist = readRankingResults(simfile)
for im, score in ranklist:
idx = im2index[im]
scores[idx, c_id] = score
makedirsforfile(resultfile)
output = open(resultfile, 'wb')
pickle.dump(
{
'concepts': concepts,
'id_images': map(int, id_images),
'scores': scores
}, output, -1)
output.close()
def precompute(self, concept):
print("[%s] precomputing candidate positive examples for %s" %
(self.name, concept))
datafile = os.path.join(self.datadir, '%s.txt' % concept)
ranklist = readRankingResults(datafile)
self.candidateset = [x[0] for x in ranklist]
self.target = concept
def _get_neighbors(self, content, context):
testCollection,testid = context.split(',')
knnfile = os.path.join(self.rootpath, testCollection, 'SimilarityIndex', testCollection, self.knndir, testid[-2:], '%s.txt' % testid)
knn = readRankingResults(knnfile)
knn = knn[:self.k]
if self.noise > 1e-3:
n = int(len(knn) * self.noise)
hits = random.sample(xrange(len(knn)), n)
random_set = random.sample(self.imset, n)
for i in range(n):
idx = hits[i]
knn[idx] = (random_set[i], 1000)
return knn
def GET(self):
input = web.input(query=None)
resp = {
'status': 0,
'hits': 0,
'random': [],
'tagrel': [],
'metric': metric,
'perf': 0
}
if input.query:
resp['status'] = 1
resp['query'] = input.query
query = input.query.lower()
if query.isdigit(): # request to view a specific image
resp['hits'] = 1
resp['tagrel'] = [{'id': query}]
return render.index(resp)
try:
names, labels = readAnnotationsFrom(collection, annotationName,
query)
name2label = dict(zip(names, labels))
except Exception, e:
name2label = {}
content = []
try:
if input.tagrel == '0':
labeled = readLabeledImageSet(collection,
query,
rootpath=rootpath)
ranklist = [(x, 0) for x in labeled]
else:
simfile = os.path.join(simdir, '%s.txt' % query)
ranklist = readRankingResults(simfile)
resp['hits'] = len(ranklist)
for name, score in ranklist:
color = 'Chartreuse' if name2label.get(name,
0) > 0 else 'red'
color = 'white' if name not in name2label else color
res = {'id': name, 'color': color}
content.append(res)
resp['perf'] = 0 if not name2label else scorer.score(
[name2label[x[0]] for x in ranklist if x[0] in name2label])
resp['tagrel'] = content[:max_hits]
except:
None
def precompute_annotator(self, concept):
INFO = 'dataengine.%s.precompute_annotator'%self.__class__.__name__
topn = 100
NegativeEngine.precompute_annotator(self, concept)
for subconcept in concept.split('-'):
expandedTagSet = set([subconcept] + wn_expand(subconcept))
try:
datafile = os.path.join(ROOT_PATH, self.collection, 'SimilarityIndex', 'ngd', '%s.txt' % subconcept)
rankedtags = readRankingResults(datafile)
expandedTagSet = expandedTagSet.union(set([x[0] for x in rankedtags[:topn]]))
except:
printError(INFO, 'failed to load ranktag file for %s' % subconcept)
self.annotator = self.annotator.union(expandedTagSet)
printStatus(INFO, 'precomputing the virtual annotator for %s: %d tags' % (concept, len(self.annotator)))
def _get_neighbors(self, content, context):
testCollection, testid = context.split(',')
knnfile = os.path.join(self.rootpath, testCollection,
'SimilarityIndex', testCollection, self.knndir,
testid[-2:], '%s.txt' % testid)
knn = readRankingResults(knnfile)
knn = knn[:self.k]
if self.noise > 1e-3:
n = int(len(knn) * self.noise)
hits = random.sample(xrange(len(knn)), n)
random_set = random.sample(self.imset, n)
for i in range(n):
idx = hits[i]
knn[idx] = (random_set[i], 1000)
return knn
def process(options, collection, annotationName, runfile, newRunName):
rootpath = options.rootpath
overwrite = options.overwrite
dataset = options.testset if options.testset else collection
concepts = readConcepts(collection, annotationName, rootpath)
simdir = os.path.join(rootpath, collection, "SimilarityIndex", dataset)
data = [x.strip() for x in open(runfile).readlines() if x.strip() and not x.strip().startswith("#")]
models = []
for line in data:
weight, run = str.split(line)
models.append((run, float(weight), 1))
for concept in concepts:
resultfile = os.path.join(simdir, newRunName, concept + ".txt")
if checkToSkip(resultfile, overwrite):
continue
scorefile = os.path.join(simdir, models[0][0], concept + ".txt")
if not os.path.exists(scorefile):
print("%s does not exist. skip" % scorefile)
continue
ranklist = readRankingResults(scorefile)
names = sorted([x[0] for x in ranklist])
nr_of_images = len(names)
name2index = dict(zip(names, range(nr_of_images)))
print("%s %d" % (concept, nr_of_images))
scoreTable = readImageScoreTable(concept, name2index, simdir, models, torank=options.torank)
assert scoreTable.shape[1] == nr_of_images
weights = [model[1] for model in models]
scores = np.matrix(weights) * scoreTable
scores = [float(scores[0, k]) for k in range(nr_of_images)]
newranklist = [(names[i], scores[i]) for i in range(nr_of_images)]
newranklist.sort(key=lambda v: (v[1], v[0]), reverse=True)
writeRankingResults(newranklist, resultfile)
def process(options, collection, annotationName, runfile, newRunName):
rootpath = options.rootpath
overwrite = options.overwrite
dataset = options.testset if options.testset else collection
concepts = readConcepts(collection, annotationName, rootpath)
simdir = os.path.join(rootpath, collection, "SimilarityIndex", dataset)
data = [x.strip() for x in open(runfile).readlines() if x.strip() and not x.strip().startswith("#")]
models = []
for line in data:
weight,run = str.split(line)
models.append((run, float(weight), 1))
for concept in concepts:
resultfile = os.path.join(simdir, newRunName, concept + ".txt")
if checkToSkip(resultfile, overwrite):
continue
scorefile = os.path.join(simdir, models[0][0], concept + ".txt")
if not os.path.exists(scorefile):
print ("%s does not exist. skip" % scorefile)
continue
ranklist = readRankingResults(scorefile)
names = sorted([x[0] for x in ranklist])
nr_of_images = len(names)
name2index = dict(zip(names, range(nr_of_images)))
print ('%s %d' % (concept, nr_of_images))
scoreTable = readImageScoreTable(concept, name2index, simdir, models, torank=options.torank)
assert(scoreTable.shape[1] == nr_of_images)
weights = [model[1] for model in models]
scores = np.matrix(weights) * scoreTable
scores = [float(scores[0,k]) for k in range(nr_of_images)]
newranklist = [(names[i], scores[i]) for i in range(nr_of_images)]
newranklist.sort(key=lambda v:(v[1],v[0]), reverse=True)
writeRankingResults(newranklist, resultfile)
def precompute_annotator(self, concept):
INFO = 'dataengine.%s.precompute_annotator' % self.__class__.__name__
topn = 100
NegativeEngine.precompute_annotator(self, concept)
for subconcept in concept.split('-'):
expandedTagSet = set([subconcept] + wn_expand(subconcept))
try:
datafile = os.path.join(ROOT_PATH, self.collection,
'SimilarityIndex', 'ngd',
'%s.txt' % subconcept)
rankedtags = readRankingResults(datafile)
expandedTagSet = expandedTagSet.union(
set([x[0] for x in rankedtags[:topn]]))
except:
printError(INFO,
'failed to load ranktag file for %s' % subconcept)
self.annotator = self.annotator.union(expandedTagSet)
printStatus(
INFO, 'precomputing the virtual annotator for %s: %d tags' %
(concept, len(self.annotator)))
def GET(self):
input = web.input(query=None)
resp = {'status':0, 'hits':0, 'random':[], 'tagrel':[], 'metric':metric, 'perf':0}
if input.query:
resp['status'] = 1
resp['query'] = input.query
query = input.query.lower()
if query.isdigit(): # request to view a specific image
resp['hits'] = 1
resp['tagrel'] = [{'id':query}]
return render.index(resp)
try:
names,labels = readAnnotationsFrom(collection, annotationName, query)
name2label = dict(zip(names,labels))
except Exception, e:
name2label = {}
content = []
try:
if input.tagrel == '0':
labeled = readLabeledImageSet(collection, query, rootpath=rootpath)
ranklist = [(x,0) for x in labeled]
else:
simfile = os.path.join(simdir, '%s.txt' % query)
ranklist = readRankingResults(simfile)
resp['hits'] = len(ranklist)
for name,score in ranklist:
color = 'Chartreuse' if name2label.get(name,0)>0 else 'red'
color = 'white' if name not in name2label else color
res = {'id':name, 'color':color}
content.append(res)
resp['perf'] = 0 if not name2label else scorer.score([name2label[x[0]] for x in ranklist if x[0] in name2label])
resp['tagrel'] = content[:max_hits]
except:
None
assert( rankMethod.startswith('tagged,lemm/%s'%collection) )
newAnnotationTemplate = annotationName[:-4] + '.' + posName + str(nr_pos) + ('.random%d'%nr_neg) + '.%d.txt'
concepts = readConcepts(collection, annotationName, rootpath)
simdir = os.path.join(rootpath, collection, 'SimilarityIndex', collection, rankMethod)
scriptfile = os.path.join(rootpath,collection,'annotationfiles', annotationName[:-4] + '.' + posName + str(nr_pos) + ('.random%d'%nr_neg) + '.0-%d.txt'%(nr_neg_bags-1))
makedirsforfile(scriptfile)
fout = open(scriptfile,'w')
fout.write('\n'.join([newAnnotationTemplate%t for t in range(nr_neg_bags)]) + '\n')
fout.close()
for concept in concepts:
simfile = os.path.join(simdir, '%s.txt' % concept)
ranklist = readRankingResults(simfile)
pos_bag = [x[0] for x in ranklist[:nr_pos]]
names, labels = readAnnotationsFrom(collection, annotationName, concept, skip_0=True, rootpath=rootpath)
negativePool = [x[0] for x in zip(names,labels) if x[1] < 0]
for t in range(nr_neg_bags):
newAnnotationName = newAnnotationTemplate % t
resultfile = os.path.join(rootpath, collection, 'Annotations', 'Image', newAnnotationName, '%s.txt'%concept)
if checkToSkip(resultfile, overwrite):
continue
true_nr_neg = max(500, len(pos_bag)*neg_pos_ratio)
neg_bag = random.sample(negativePool, true_nr_neg) #len(pos_bag)*neg_pos_ratio)
assert(len(set(pos_bag).intersection(set(neg_bag))) == 0)
printStatus(INFO, "anno(%s,%d) %d pos %d neg -> %s" % (concept,t,len(pos_bag),len(neg_bag),resultfile))
writeAnnotations(pos_bag + neg_bag, [1]*len(pos_bag) + [-1]*len(neg_bag), resultfile)
#posName = 'ccgd' + str(numPos)
#tagrelMethod = 'flickr1m/ccgd,knn,1000'
concepts = readConcepts(collection, sourceAnnotationName % 0, rootpath)
holdoutfile = os.path.join(rootpath, collection, "ImageSets",
"holdout.txt")
holdoutSet = set(map(str.strip, open(holdoutfile).readlines()))
print('%s holdout %d' % (collection, len(holdoutSet)))
for concept in concepts:
simfile = os.path.join(rootpath, collection, 'SimilarityIndex',
collection, 'tagged,lemm', tagrelMethod,
'%s.txt' % concept)
searchresults = readRankingResults(simfile)
searchresults = [x for x in searchresults if x[0] not in holdoutSet]
positiveSet = [x[0] for x in searchresults[:numPos]]
for t in range(T):
newAnnotationName = sourceAnnotationName % t
newAnnotationName = newAnnotationName.replace(
'rand%d.0' % numPos, posName)
names, labels = readAnnotationsFrom(collection,
sourceAnnotationName % t,
concept, rootpath)
negativeSet = [x[0] for x in zip(names, labels) if -1 == x[1]]
renamed = positiveSet + negativeSet
relabeled = [1] * len(positiveSet) + [-1] * len(negativeSet)
print('[%s] %s +%d, -%d -> %s' %
#tagrelMethod = 'textual'
posName = 'clickcount' + str(numPos)
tagrelMethod = 'clickcount'
#posName = 'ccgd' + str(numPos)
#tagrelMethod = 'flickr1m/ccgd,knn,1000'
concepts = readConcepts(collection, sourceAnnotationName%0, rootpath)
holdoutfile = os.path.join(rootpath, collection, "ImageSets", "holdout.txt")
holdoutSet = set(map(str.strip, open(holdoutfile).readlines()))
print ('%s holdout %d' % (collection,len(holdoutSet)))
for concept in concepts:
simfile = os.path.join(rootpath, collection, 'SimilarityIndex', collection, 'tagged,lemm', tagrelMethod, '%s.txt' % concept)
searchresults = readRankingResults(simfile)
searchresults = [x for x in searchresults if x[0] not in holdoutSet]
positiveSet = [x[0] for x in searchresults[:numPos]]
for t in range(T):
newAnnotationName = sourceAnnotationName % t
newAnnotationName = newAnnotationName.replace('rand%d.0'%numPos, posName)
names,labels = readAnnotationsFrom(collection,sourceAnnotationName%t,concept,rootpath)
negativeSet = [x[0] for x in zip(names,labels) if -1 == x[1]]
renamed = positiveSet + negativeSet
relabeled = [1] * len(positiveSet) + [-1] * len(negativeSet)
print ('[%s] %s +%d, -%d -> %s' % (concept,sourceAnnotationName % t,len(positiveSet),len(negativeSet),newAnnotationName))
writeAnnotationsTo(renamed, relabeled, collection, newAnnotationName, concept, rootpath)
for t in range(T):
rankMethod)
scriptfile = os.path.join(
rootpath, collection, 'annotationfiles',
annotationName[:-4] + '.' + posName + str(nr_pos) +
('.random%d' % nr_neg) + '.0-%d.txt' % (nr_neg_bags - 1))
makedirsforfile(scriptfile)
fout = open(scriptfile, 'w')
fout.write(
'\n'.join([newAnnotationTemplate % t
for t in range(nr_neg_bags)]) + '\n')
fout.close()
for concept in concepts:
simfile = os.path.join(simdir, '%s.txt' % concept)
ranklist = readRankingResults(simfile)
pos_bag = [x[0] for x in ranklist[:nr_pos]]
names, labels = readAnnotationsFrom(collection,
annotationName,
concept,
skip_0=True,
rootpath=rootpath)
negativePool = [x[0] for x in zip(names, labels) if x[1] < 0]
for t in range(nr_neg_bags):
newAnnotationName = newAnnotationTemplate % t
resultfile = os.path.join(rootpath, collection, 'Annotations',
'Image', newAnnotationName,
'%s.txt' % concept)
if checkToSkip(resultfile, overwrite):
continue
def precompute(self, concept):
print ("[%s] precomputing candidate positive examples for %s" % (self.name, concept))
datafile = os.path.join(self.datadir, '%s.txt' % concept)
ranklist = readRankingResults(datafile)
self.candidateset = [x[0] for x in ranklist]
self.target = concept