redefining the scope of SIG AI

sig-ai/charter.md

@@ -4,29 +4,29 @@ This charter adheres to the conventions described in [KubeEdge Open Governance](
 
 ## Scope
 
-SIG AI is responsible for providing general platform capabilities based on KubeEdge so that AI applications running at the edge can benefit from cost reduction, model performance improvement, and data privacy protection.
+The SIG AI focuses on technical discussion, API definition, reference architecture, implementation in the Edge AI field, to enable AI applications better running on edge (including cost saving, performance improvement, and data protection).
     
 ### In scope
 #### Areas of Focus
 1. Empower KubeEdge with existing AI ecosystems, to support execution of Edge AI applications and services：
     - Support heterogeneous edge hardware, e.g., Ascend, Kunlun, Cambrian, and Rockchip
-    - Integrate typical AI frameworks into KubeEdge, e.g., Tensorflow, Pytorch, PaddlePaddle and Mindspore 
+    - Integrate typical AI frameworks into KubeEdge, e.g., Tensorflow, Pytorch, PaddlePaddle and Mindspore etc.
     - Integrate KubeFlow and ONNX into KubeEdge, to enable interoperability of edge models with diverse formats
     - Cooperate with other open source communities, e.g., Akraino and LF AI
 1. Research **includes but not limited to**:
-    - cloud training and edge inference
-    - incremental learning
-    - edge-cloud-collaborative inference
-        - knowledge distillation for the cloud and edge model
-    - federated learning 
-    - edge model and dataset management
-1. Furnish an end-to-end Edge AI benchmarking framework, to identify best practices of Edge AI applications-and-services development:
+    - Cloud training and edge inference
+    - Incremental learning
+    - Edge-cloud-collaborative inference
+        - Knowledge distillation for the cloud and edge model
+    - Federated learning 
+    - Edge model and dataset management
+1. Edge AI benchmarking relevant work, to identify most important dimensions when developing, evaluating Edge AI applications-and-services system:
     - Provide Contextual Metrics
-        - for typical Edge AI applications scenarios
+        - For typical Edge AI applications scenarios
     - Provide Standardized Evaluation Settings
-        - standardized datasets, architectures, and hardware
-        - for each routine AI module, e.g., data collection, data preprocessing, train and inference 
-        - for each architecture layer, i.e., cloud, edge, and end-device 
+        - Standardized datasets, architectures, and hardware
+        - For each routine AI module, e.g., data collection, data preprocessing, train and inference 
+        - For each architecture layer, i.e., cloud, edge, and end-device 
 ### Out of scope
 - Re-invent existing AI framework, e.g., Tensorflow, Pytorch and Mindspore
 - Offer domain/application-specific algorithms, e.g., facial recognition and text classification