Amazon Rekognition - Python Code Samples

amazon-rekognition.md

Amazon Rekognition - Python Code Samples

1. Labels Detection
2. Faces Detection
3. Faces Comparison
4. Faces Indexing
5. Faces Search

test-1-detect-labels.py

import boto3

BUCKET = "amazon-rekognition"
KEY = "test.jpg"

def detect_labels(bucket, key, max_labels=10, min_confidence=90, region="eu-west-1"):
	rekognition = boto3.client("rekognition", region)
	response = rekognition.detect_labels(
		Image={
			"S3Object": {
				"Bucket": bucket,
				"Name": key,
			}
		},
		MaxLabels=max_labels,
		MinConfidence=min_confidence,
	)
	return response['Labels']


for label in detect_labels(BUCKET, KEY):
	print "{Name} - {Confidence}%".format(**label)

"""
	Expected output:

	People - 99.2436447144%
	Person - 99.2436447144%
	Human - 99.2351226807%
	Clothing - 96.7797698975%
	Suit - 96.7797698975%

"""

test-2-detect-faces.py

import boto3

BUCKET = "amazon-rekognition"
KEY = "test.jpg"
FEATURES_BLACKLIST = ("Landmarks", "Emotions", "Pose", "Quality", "BoundingBox", "Confidence")

def detect_faces(bucket, key, attributes=['ALL'], region="eu-west-1"):
	rekognition = boto3.client("rekognition", region)
	response = rekognition.detect_faces(
	    Image={
			"S3Object": {
				"Bucket": bucket,
				"Name": key,
			}
		},
	    Attributes=attributes,
	)
	return response['FaceDetails']

for face in detect_faces(BUCKET, KEY):
	print "Face ({Confidence}%)".format(**face)
	# emotions
	for emotion in face['Emotions']:
		print "  {Type} : {Confidence}%".format(**emotion)
	# quality
	for quality, value in face['Quality'].iteritems():
		print "  {quality} : {value}".format(quality=quality, value=value)
	# facial features
	for feature, data in face.iteritems():
		if feature not in FEATURES_BLACKLIST:
			print "  {feature}({data[Value]}) : {data[Confidence]}%".format(feature=feature, data=data)

"""
	Expected output:

	Face (99.945602417%)
	  SAD : 14.6038293839%
	  HAPPY : 12.3668470383%
	  DISGUSTED : 3.81404161453%
	  Sharpness : 10.0
	  Brightness : 31.4071826935
	  Eyeglasses(False) : 99.990234375%
	  Sunglasses(False) : 99.9500656128%
	  Gender(Male) : 99.9291687012%
	  EyesOpen(True) : 99.9609146118%
	  Smile(False) : 99.8329467773%
	  MouthOpen(False) : 98.3746566772%
	  Mustache(False) : 98.7549591064%
	  Beard(False) : 92.758682251%

"""

test-3-compare-faces.py

import boto3

BUCKET = "amazon-rekognition"
KEY_SOURCE = "test.jpg"
KEY_TARGET = "target.jpg"

def compare_faces(bucket, key, bucket_target, key_target, threshold=80, region="eu-west-1"):
	rekognition = boto3.client("rekognition", region)
	response = rekognition.compare_faces(
	    SourceImage={
			"S3Object": {
				"Bucket": bucket,
				"Name": key,
			}
		},
		TargetImage={
			"S3Object": {
				"Bucket": bucket_target,
				"Name": key_target,
			}
		},
	    SimilarityThreshold=threshold,
	)
	return response['SourceImageFace'], response['FaceMatches']


source_face, matches = compare_faces(BUCKET, KEY_SOURCE, BUCKET, KEY_TARGET)

# the main source face
print "Source Face ({Confidence}%)".format(**source_face)

# one match for each target face
for match in matches:
	print "Target Face ({Confidence}%)".format(**match['Face'])
	print "  Similarity : {}%".format(match['Similarity'])

"""
    Expected output:
    
    Source Face (99.945602417%)
    Target Face (99.9963378906%)
      Similarity : 89.0%

"""

test-4-index-faces.py

import boto3

BUCKET = "amazon-rekognition"
KEY = "test.jpg"
IMAGE_ID = KEY  # S3 key as ImageId
COLLECTION = "my-collection-id"

# Note: you have to create the collection first!
# rekognition.create_collection(CollectionId=COLLECTION)

def index_faces(bucket, key, collection_id, image_id=None, attributes=(), region="eu-west-1"):
	rekognition = boto3.client("rekognition", region)
	response = rekognition.index_faces(
		Image={
			"S3Object": {
				"Bucket": bucket,
				"Name": key,
			}
		},
		CollectionId=collection_id,
		ExternalImageId=image_id,
	    DetectionAttributes=attributes,
	)
	return response['FaceRecords']


for record in index_faces(BUCKET, KEY, COLLECTION, IMAGE_ID):
	face = record['Face']
	# details = record['FaceDetail']
	print "Face ({}%)".format(face['Confidence'])
	print "  FaceId: {}".format(face['FaceId'])
	print "  ImageId: {}".format(face['ImageId'])


"""
	Expected output:

	Face (99.945602417%)
	  FaceId: dc090f86-48a4-5f09-905f-44e97fb1d455
	  ImageId: f974c8d3-7519-5796-a08d-b96e0f2fc242

"""

test-5-search-faces.py

import boto3

BUCKET = "amazon-rekognition"
KEY = "search.jpg"
COLLECTION = "my-collection-id"

def search_faces_by_image(bucket, key, collection_id, threshold=80, region="eu-west-1"):
	rekognition = boto3.client("rekognition", region)
	response = rekognition.search_faces_by_image(
		Image={
			"S3Object": {
				"Bucket": bucket,
				"Name": key,
			}
		},
		CollectionId=collection_id,
		FaceMatchThreshold=threshold,
	)
	return response['FaceMatches']

for record in search_faces_by_image(BUCKET, KEY, COLLECTION):
	face = record['Face']
	print "Matched Face ({}%)".format(record['Similarity'])
	print "  FaceId : {}".format(face['FaceId'])
	print "  ImageId : {}".format(face['ExternalImageId'])


"""
	Expected output:

	Matched Face (96.6647949219%)
	  FaceId : dc090f86-48a4-5f09-905f-44e97fb1d455
	  ImageId : test.jpg

"""

@toto88 you can just remove them from the labels list in your favorite programming language (if you are using one of the available AWS SDK). The detectLabel response is just a JSON object so you should be able to parse it with any JSON parser out there, exclude the "Person"/"People"/"Human" labels and use/store all the others.

@alexcasalboni thanks for your reply, how i can get the full list of labels, I have checked the Q&A page, it's just have some examples. https://aws.amazon.com/rekognition/faqs/

@toto88 unfortunately there is no public list of labels available yet.. Rekognition uses thousands of labels and the model is updated periodically under the hood.. for now, the best way to know if a given label is recognized is to submit a few images with that object/entity and verify it directly :)

@alexcasalboni thanks for all your effort! I tried your script but having an issue with print. Do you have any idea what I can do? Also, MaxLables and MinConfidence does not seem to work either.

@sunhoro I bet you're using Python3 :) You just need to convert the script to Python 3 (hint: use the print function, with parenthesis).

The face emotion part displays all the types of available emotions. How can I display the emotion with highest confidence?

Just parse through the json like output that it gives using the dictionary key for what you want to display. I don’t remember the output exactly right now but a simple for loop and key pair logic should get you what you want.

…

On Thu, Oct 1, 2020 at 10:43 AM AmarNaga ***@***.***> wrote: ***@***.**** commented on this gist. ------------------------------ The face emotion part displays all the types of available emotions. How can I display the emotion with highest confidence? — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <https://urldefense.proofpoint.com/v2/url?u=https-3A__gist.github.com_0f21a1889f09760f8981b643326730ff-23gistcomment-2D3473971&d=DwMCaQ&c=slrrB7dE8n7gBJbeO0g-IQ&r=r_isl-LMqNv6X9xB1i6A3Q&m=rlwMO1NLOWcCFcM-3bdAjQP6q3cV17vnemJH1HsVCMk&s=UOP9jadqSjtxJbsKMWHoW01yqkdlvHyswun_SVrbcrI&e=>, or unsubscribe <https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe-2Dauth_AKWZQB4RYVU4ZSQPNMAUKNDSISIQPANCNFSM4HYVRF3Q&d=DwMCaQ&c=slrrB7dE8n7gBJbeO0g-IQ&r=r_isl-LMqNv6X9xB1i6A3Q&m=rlwMO1NLOWcCFcM-3bdAjQP6q3cV17vnemJH1HsVCMk&s=uXsXVCd0la58_7bgkxccN-bapUw6DDDNITPn2GusKp8&e=> .