Observed since 20 Oct, 17:57 UTC
No superseded metadata states collected. Unchanged observations do not create new revisions.
access-client
# Access Client Python client to interact with acessai's various applications/services like access-face-vision ### Installation ```bash pip install access-cleint ``` ### Usage ```python from access_client.face_vision.client import AFV #instantiate by giving the host and port of the access-face-vision server afv = AFV("http://localhost:5001") ``` Now we need to create a face-group. This is will be our face-index. ```python afv.create_face_group(face_group_name="celebrities") #Pass in the directory and face group name afv.add_faces_to_face_group(dir="./samples/celebrities", face_group="celebrities") # Directory structure # **/Images/ # A/ # A_K_01.jpg # A_K_02.jpg # B/ # B_S_01.jpg # B_S_02.jp
Python client to interact with acessai's various applications/services like access-face-vision
pip install access-cleint
from access_client.face_vision.client import AFV
#instantiate by giving the host and port of the access-face-vision server
afv = AFV("http://localhost:5001")
Now we need to create a face-group. This is will be our face-index.
afv.create_face_group(face_group_name="celebrities")
#Pass in the directory and face group name
afv.add_faces_to_face_group(dir="./samples/celebrities", face_group="celebrities")
# Directory structure
# **/Images/
# A/
# A_K_01.jpg
# A_K_02.jpg
# B/
# B_S_01.jpg
# B_S_02.jpg
Once faces are indexed we can run inference on it.
afv.parse(img_path="./samples/celebrities/Bill Gates/Bill_Gates.jpg", face_group="celebrities")
Download counts and popularity will appear here as the collector receives them.
Declarations describe requirements; they are not a resolved dependency graph.
sdist · 7.15 KB · Python source
Observed since 20 Oct, 17:57 UTC
No superseded metadata states collected. Unchanged observations do not create new revisions.
package_version:cd5f13a7236f4dd537be68adeb35840bac555fee9bb3baf4acaabe921b2fde23