// Copyright 2020 Google LLC
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
//     http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.

syntax = "proto3";

package google.cloud.automl.v1beta1;

import "google/api/annotations.proto";
import "google/api/client.proto";
import "google/api/field_behavior.proto";
import "google/api/resource.proto";
import "google/cloud/automl/v1beta1/annotation_payload.proto";
import "google/cloud/automl/v1beta1/data_items.proto";
import "google/cloud/automl/v1beta1/io.proto";
import "google/cloud/automl/v1beta1/operations.proto";
import "google/longrunning/operations.proto";

option go_package = "cloud.google.com/go/automl/apiv1beta1/automlpb;automlpb";
option java_multiple_files = true;
option java_outer_classname = "PredictionServiceProto";
option java_package = "com.google.cloud.automl.v1beta1";
option php_namespace = "Google\\Cloud\\AutoMl\\V1beta1";
option ruby_package = "Google::Cloud::AutoML::V1beta1";

// AutoML Prediction API.
//
// On any input that is documented to expect a string parameter in
// snake_case or kebab-case, either of those cases is accepted.
service PredictionService {
  option (google.api.default_host) = "automl.googleapis.com";
  option (google.api.oauth_scopes) = "https://www.googleapis.com/auth/cloud-platform";

  // Perform an online prediction. The prediction result will be directly
  // returned in the response.
  // Available for following ML problems, and their expected request payloads:
  // * Image Classification - Image in .JPEG, .GIF or .PNG format, image_bytes
  //                          up to 30MB.
  // * Image Object Detection - Image in .JPEG, .GIF or .PNG format, image_bytes
  //                            up to 30MB.
  // * Text Classification - TextSnippet, content up to 60,000 characters,
  //                         UTF-8 encoded.
  // * Text Extraction - TextSnippet, content up to 30,000 characters,
  //                     UTF-8 NFC encoded.
  // * Translation - TextSnippet, content up to 25,000 characters, UTF-8
  //                 encoded.
  // * Tables - Row, with column values matching the columns of the model,
  //            up to 5MB. Not available for FORECASTING
  //
  // [prediction_type][google.cloud.automl.v1beta1.TablesModelMetadata.prediction_type].
  // * Text Sentiment - TextSnippet, content up 500 characters, UTF-8
  //                     encoded.
  rpc Predict(PredictRequest) returns (PredictResponse) {
    option (google.api.http) = {
      post: "/v1beta1/{name=projects/*/locations/*/models/*}:predict"
      body: "*"
    };
    option (google.api.method_signature) = "name,payload,params";
  }

  // Perform a batch prediction. Unlike the online [Predict][google.cloud.automl.v1beta1.PredictionService.Predict], batch
  // prediction result won't be immediately available in the response. Instead,
  // a long running operation object is returned. User can poll the operation
  // result via [GetOperation][google.longrunning.Operations.GetOperation]
  // method. Once the operation is done, [BatchPredictResult][google.cloud.automl.v1beta1.BatchPredictResult] is returned in
  // the [response][google.longrunning.Operation.response] field.
  // Available for following ML problems:
  // * Image Classification
  // * Image Object Detection
  // * Video Classification
  // * Video Object Tracking * Text Extraction
  // * Tables
  rpc BatchPredict(BatchPredictRequest) returns (google.longrunning.Operation) {
    option (google.api.http) = {
      post: "/v1beta1/{name=projects/*/locations/*/models/*}:batchPredict"
      body: "*"
    };
    option (google.api.method_signature) = "name,input_config,output_config,params";
    option (google.longrunning.operation_info) = {
      response_type: "BatchPredictResult"
      metadata_type: "OperationMetadata"
    };
  }
}

// Request message for [PredictionService.Predict][google.cloud.automl.v1beta1.PredictionService.Predict].
message PredictRequest {
  // Required. Name of the model requested to serve the prediction.
  string name = 1 [
    (google.api.field_behavior) = REQUIRED,
    (google.api.resource_reference) = {
      type: "automl.googleapis.com/Model"
    }
  ];

  // Required. Payload to perform a prediction on. The payload must match the
  // problem type that the model was trained to solve.
  ExamplePayload payload = 2 [(google.api.field_behavior) = REQUIRED];

  // Additional domain-specific parameters, any string must be up to 25000
  // characters long.
  //
  // *  For Image Classification:
  //
  //    `score_threshold` - (float) A value from 0.0 to 1.0. When the model
  //     makes predictions for an image, it will only produce results that have
  //     at least this confidence score. The default is 0.5.
  //
  //  *  For Image Object Detection:
  //    `score_threshold` - (float) When Model detects objects on the image,
  //        it will only produce bounding boxes which have at least this
  //        confidence score. Value in 0 to 1 range, default is 0.5.
  //    `max_bounding_box_count` - (int64) No more than this number of bounding
  //        boxes will be returned in the response. Default is 100, the
  //        requested value may be limited by server.
  // *  For Tables:
  //    feature_imp<span>ortan</span>ce - (boolean) Whether feature importance
  //        should be populated in the returned TablesAnnotation.
  //        The default is false.
  map<string, string> params = 3;
}

// Response message for [PredictionService.Predict][google.cloud.automl.v1beta1.PredictionService.Predict].
message PredictResponse {
  // Prediction result.
  // Translation and Text Sentiment will return precisely one payload.
  repeated AnnotationPayload payload = 1;

  // The preprocessed example that AutoML actually makes prediction on.
  // Empty if AutoML does not preprocess the input example.
  // * For Text Extraction:
  //   If the input is a .pdf file, the OCR'ed text will be provided in
  //   [document_text][google.cloud.automl.v1beta1.Document.document_text].
  ExamplePayload preprocessed_input = 3;

  // Additional domain-specific prediction response metadata.
  //
  // * For Image Object Detection:
  //  `max_bounding_box_count` - (int64) At most that many bounding boxes per
  //      image could have been returned.
  //
  // * For Text Sentiment:
  //  `sentiment_score` - (float, deprecated) A value between -1 and 1,
  //      -1 maps to least positive sentiment, while 1 maps to the most positive
  //      one and the higher the score, the more positive the sentiment in the
  //      document is. Yet these values are relative to the training data, so
  //      e.g. if all data was positive then -1 will be also positive (though
  //      the least).
  //      The sentiment_score shouldn't be confused with "score" or "magnitude"
  //      from the previous Natural Language Sentiment Analysis API.
  map<string, string> metadata = 2;
}

// Request message for [PredictionService.BatchPredict][google.cloud.automl.v1beta1.PredictionService.BatchPredict].
message BatchPredictRequest {
  // Required. Name of the model requested to serve the batch prediction.
  string name = 1 [
    (google.api.field_behavior) = REQUIRED,
    (google.api.resource_reference) = {
      type: "automl.googleapis.com/Model"
    }
  ];

  // Required. The input configuration for batch prediction.
  BatchPredictInputConfig input_config = 3 [(google.api.field_behavior) = REQUIRED];

  // Required. The Configuration specifying where output predictions should
  // be written.
  BatchPredictOutputConfig output_config = 4 [(google.api.field_behavior) = REQUIRED];

  // Required. Additional domain-specific parameters for the predictions, any string must
  // be up to 25000 characters long.
  //
  // *  For Text Classification:
  //
  //    `score_threshold` - (float) A value from 0.0 to 1.0. When the model
  //         makes predictions for a text snippet, it will only produce results
  //         that have at least this confidence score. The default is 0.5.
  //
  // *  For Image Classification:
  //
  //    `score_threshold` - (float) A value from 0.0 to 1.0. When the model
  //         makes predictions for an image, it will only produce results that
  //         have at least this confidence score. The default is 0.5.
  //
  // *  For Image Object Detection:
  //
  //    `score_threshold` - (float) When Model detects objects on the image,
  //        it will only produce bounding boxes which have at least this
  //        confidence score. Value in 0 to 1 range, default is 0.5.
  //    `max_bounding_box_count` - (int64) No more than this number of bounding
  //        boxes will be produced per image. Default is 100, the
  //        requested value may be limited by server.
  //
  // *  For Video Classification :
  //
  //    `score_threshold` - (float) A value from 0.0 to 1.0. When the model
  //        makes predictions for a video, it will only produce results that
  //        have at least this confidence score. The default is 0.5.
  //    `segment_classification` - (boolean) Set to true to request
  //        segment-level classification. AutoML Video Intelligence returns
  //        labels and their confidence scores for the entire segment of the
  //        video that user specified in the request configuration.
  //        The default is "true".
  //    `shot_classification` - (boolean) Set to true to request shot-level
  //        classification. AutoML Video Intelligence determines the boundaries
  //        for each camera shot in the entire segment of the video that user
  //        specified in the request configuration. AutoML Video Intelligence
  //        then returns labels and their confidence scores for each detected
  //        shot, along with the start and end time of the shot.
  //        WARNING: Model evaluation is not done for this classification type,
  //        the quality of it depends on training data, but there are no metrics
  //        provided to describe that quality. The default is "false".
  //    `1s_interval_classification` - (boolean) Set to true to request
  //        classification for a video at one-second intervals. AutoML Video
  //        Intelligence returns labels and their confidence scores for each
  //        second of the entire segment of the video that user specified in the
  //        request configuration.
  //        WARNING: Model evaluation is not done for this classification
  //        type, the quality of it depends on training data, but there are no
  //        metrics provided to describe that quality. The default is
  //        "false".
  //
  // *  For Tables:
  //
  //    feature_imp<span>ortan</span>ce - (boolean) Whether feature importance
  //        should be populated in the returned TablesAnnotations. The
  //        default is false.
  //
  // *  For Video Object Tracking:
  //
  //    `score_threshold` - (float) When Model detects objects on video frames,
  //        it will only produce bounding boxes which have at least this
  //        confidence score. Value in 0 to 1 range, default is 0.5.
  //    `max_bounding_box_count` - (int64) No more than this number of bounding
  //        boxes will be returned per frame. Default is 100, the requested
  //        value may be limited by server.
  //    `min_bounding_box_size` - (float) Only bounding boxes with shortest edge
  //      at least that long as a relative value of video frame size will be
  //      returned. Value in 0 to 1 range. Default is 0.
  map<string, string> params = 5 [(google.api.field_behavior) = REQUIRED];
}

// Result of the Batch Predict. This message is returned in
// [response][google.longrunning.Operation.response] of the operation returned
// by the [PredictionService.BatchPredict][google.cloud.automl.v1beta1.PredictionService.BatchPredict].
message BatchPredictResult {
  // Additional domain-specific prediction response metadata.
  //
  // *  For Image Object Detection:
  //  `max_bounding_box_count` - (int64) At most that many bounding boxes per
  //      image could have been returned.
  //
  // *  For Video Object Tracking:
  //  `max_bounding_box_count` - (int64) At most that many bounding boxes per
  //      frame could have been returned.
  map<string, string> metadata = 1;
}