// Copyright 2017 Google Inc.
//
// 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.videointelligence.v1beta1;

import "google/api/annotations.proto";
import "google/longrunning/operations.proto";
import "google/protobuf/timestamp.proto";
import "google/rpc/status.proto";

option csharp_namespace = "Google.Cloud.VideoIntelligence.V1Beta1";
option go_package = "google.golang.org/genproto/googleapis/cloud/videointelligence/v1beta1;videointelligence";
option java_multiple_files = true;
option java_outer_classname = "VideoIntelligenceServiceProto";
option java_package = "com.google.cloud.videointelligence.v1beta1";


// Service that implements Google Cloud Video Intelligence API.
service VideoIntelligenceService {
  // Performs asynchronous video annotation. Progress and results can be
  // retrieved through the `google.longrunning.Operations` interface.
  // `Operation.metadata` contains `AnnotateVideoProgress` (progress).
  // `Operation.response` contains `AnnotateVideoResponse` (results).
  rpc AnnotateVideo(AnnotateVideoRequest) returns (google.longrunning.Operation) {
    option (google.api.http) = { post: "/v1beta1/videos:annotate" body: "*" };
  }
}

// Video annotation request.
message AnnotateVideoRequest {
  // Input video location. Currently, only
  // [Google Cloud Storage](https://cloud.google.com/storage/) URIs are
  // supported, which must be specified in the following format:
  // `gs://bucket-id/object-id` (other URI formats return
  // [google.rpc.Code.INVALID_ARGUMENT][google.rpc.Code.INVALID_ARGUMENT]). For more information, see
  // [Request URIs](/storage/docs/reference-uris).
  // A video URI may include wildcards in `object-id`, and thus identify
  // multiple videos. Supported wildcards: '*' to match 0 or more characters;
  // '?' to match 1 character. If unset, the input video should be embedded
  // in the request as `input_content`. If set, `input_content` should be unset.
  string input_uri = 1;

  // The video data bytes. Encoding: base64. If unset, the input video(s)
  // should be specified via `input_uri`. If set, `input_uri` should be unset.
  string input_content = 6;

  // Requested video annotation features.
  repeated Feature features = 2;

  // Additional video context and/or feature-specific parameters.
  VideoContext video_context = 3;

  // Optional location where the output (in JSON format) should be stored.
  // Currently, only [Google Cloud Storage](https://cloud.google.com/storage/)
  // URIs are supported, which must be specified in the following format:
  // `gs://bucket-id/object-id` (other URI formats return
  // [google.rpc.Code.INVALID_ARGUMENT][google.rpc.Code.INVALID_ARGUMENT]). For more information, see
  // [Request URIs](/storage/docs/reference-uris).
  string output_uri = 4;

  // Optional cloud region where annotation should take place. Supported cloud
  // regions: `us-east1`, `us-west1`, `europe-west1`, `asia-east1`. If no region
  // is specified, a region will be determined based on video file location.
  string location_id = 5;
}

// Video context and/or feature-specific parameters.
message VideoContext {
  // Video segments to annotate. The segments may overlap and are not required
  // to be contiguous or span the whole video. If unspecified, each video
  // is treated as a single segment.
  repeated VideoSegment segments = 1;

  // If label detection has been requested, what labels should be detected
  // in addition to video-level labels or segment-level labels. If unspecified,
  // defaults to `SHOT_MODE`.
  LabelDetectionMode label_detection_mode = 2;

  // Whether the video has been shot from a stationary (i.e. non-moving) camera.
  // When set to true, might improve detection accuracy for moving objects.
  bool stationary_camera = 3;

  // Model to use for label detection.
  // Supported values: "latest" and "stable" (the default).
  string label_detection_model = 4;

  // Model to use for face detection.
  // Supported values: "latest" and "stable" (the default).
  string face_detection_model = 5;

  // Model to use for shot change detection.
  // Supported values: "latest" and "stable" (the default).
  string shot_change_detection_model = 6;

  // Model to use for safe search detection.
  // Supported values: "latest" and "stable" (the default).
  string safe_search_detection_model = 7;
}

// Video segment.
message VideoSegment {
  // Start offset in microseconds (inclusive). Unset means 0.
  int64 start_time_offset = 1;

  // End offset in microseconds (inclusive). Unset means 0.
  int64 end_time_offset = 2;
}

// Label location.
message LabelLocation {
  // Video segment. Set to [-1, -1] for video-level labels.
  // Set to [timestamp, timestamp] for frame-level labels.
  // Otherwise, corresponds to one of `AnnotateSpec.segments`
  // (if specified) or to shot boundaries (if requested).
  VideoSegment segment = 1;

  // Confidence that the label is accurate. Range: [0, 1].
  float confidence = 2;

  // Label level.
  LabelLevel level = 3;
}

// Label annotation.
message LabelAnnotation {
  // Textual description, e.g. `Fixed-gear bicycle`.
  string description = 1;

  // Language code for `description` in BCP-47 format.
  string language_code = 2;

  // Where the label was detected and with what confidence.
  repeated LabelLocation locations = 3;
}

// Safe search annotation (based on per-frame visual signals only).
// If no unsafe content has been detected in a frame, no annotations
// are present for that frame. If only some types of unsafe content
// have been detected in a frame, the likelihood is set to `UNKNOWN`
// for all other types of unsafe content.
message SafeSearchAnnotation {
  // Likelihood of adult content.
  Likelihood adult = 1;

  // Likelihood that an obvious modification was made to the original
  // version to make it appear funny or offensive.
  Likelihood spoof = 2;

  // Likelihood of medical content.
  Likelihood medical = 3;

  // Likelihood of violent content.
  Likelihood violent = 4;

  // Likelihood of racy content.
  Likelihood racy = 5;

  // Video time offset in microseconds.
  int64 time_offset = 6;
}

// Bounding box.
message BoundingBox {
  // Left X coordinate.
  int32 left = 1;

  // Right X coordinate.
  int32 right = 2;

  // Bottom Y coordinate.
  int32 bottom = 3;

  // Top Y coordinate.
  int32 top = 4;
}

// Face location.
message FaceLocation {
  // Bounding box in a frame.
  BoundingBox bounding_box = 1;

  // Video time offset in microseconds.
  int64 time_offset = 2;
}

// Face annotation.
message FaceAnnotation {
  // Thumbnail of a representative face view (in JPEG format). Encoding: base64.
  string thumbnail = 1;

  // All locations where a face was detected.
  // Faces are detected and tracked on a per-video basis
  // (as opposed to across multiple videos).
  repeated VideoSegment segments = 2;

  // Face locations at one frame per second.
  repeated FaceLocation locations = 3;
}

// Annotation results for a single video.
message VideoAnnotationResults {
  // Video file location in
  // [Google Cloud Storage](https://cloud.google.com/storage/).
  string input_uri = 1;

  // Label annotations. There is exactly one element for each unique label.
  repeated LabelAnnotation label_annotations = 2;

  // Face annotations. There is exactly one element for each unique face.
  repeated FaceAnnotation face_annotations = 3;

  // Shot annotations. Each shot is represented as a video segment.
  repeated VideoSegment shot_annotations = 4;

  // Safe search annotations.
  repeated SafeSearchAnnotation safe_search_annotations = 6;

  // If set, indicates an error. Note that for a single `AnnotateVideoRequest`
  // some videos may succeed and some may fail.
  google.rpc.Status error = 5;
}

// Video annotation response. Included in the `response`
// field of the `Operation` returned by the `GetOperation`
// call of the `google::longrunning::Operations` service.
message AnnotateVideoResponse {
  // Annotation results for all videos specified in `AnnotateVideoRequest`.
  repeated VideoAnnotationResults annotation_results = 1;
}

// Annotation progress for a single video.
message VideoAnnotationProgress {
  // Video file location in
  // [Google Cloud Storage](https://cloud.google.com/storage/).
  string input_uri = 1;

  // Approximate percentage processed thus far.
  // Guaranteed to be 100 when fully processed.
  int32 progress_percent = 2;

  // Time when the request was received.
  google.protobuf.Timestamp start_time = 3;

  // Time of the most recent update.
  google.protobuf.Timestamp update_time = 4;
}

// Video annotation progress. Included in the `metadata`
// field of the `Operation` returned by the `GetOperation`
// call of the `google::longrunning::Operations` service.
message AnnotateVideoProgress {
  // Progress metadata for all videos specified in `AnnotateVideoRequest`.
  repeated VideoAnnotationProgress annotation_progress = 1;
}

// Video annotation feature.
enum Feature {
  // Unspecified.
  FEATURE_UNSPECIFIED = 0;

  // Label detection. Detect objects, such as dog or flower.
  LABEL_DETECTION = 1;

  // Human face detection and tracking.
  FACE_DETECTION = 2;

  // Shot change detection.
  SHOT_CHANGE_DETECTION = 3;

  // Safe search detection.
  SAFE_SEARCH_DETECTION = 4;
}

// Label level (scope).
enum LabelLevel {
  // Unspecified.
  LABEL_LEVEL_UNSPECIFIED = 0;

  // Video-level. Corresponds to the whole video.
  VIDEO_LEVEL = 1;

  // Segment-level. Corresponds to one of `AnnotateSpec.segments`.
  SEGMENT_LEVEL = 2;

  // Shot-level. Corresponds to a single shot (i.e. a series of frames
  // without a major camera position or background change).
  SHOT_LEVEL = 3;

  // Frame-level. Corresponds to a single video frame.
  FRAME_LEVEL = 4;
}

// Label detection mode.
enum LabelDetectionMode {
  // Unspecified.
  LABEL_DETECTION_MODE_UNSPECIFIED = 0;

  // Detect shot-level labels.
  SHOT_MODE = 1;

  // Detect frame-level labels.
  FRAME_MODE = 2;

  // Detect both shot-level and frame-level labels.
  SHOT_AND_FRAME_MODE = 3;
}

// Bucketized representation of likelihood.
enum Likelihood {
  // Unknown likelihood.
  UNKNOWN = 0;

  // Very unlikely.
  VERY_UNLIKELY = 1;

  // Unlikely.
  UNLIKELY = 2;

  // Possible.
  POSSIBLE = 3;

  // Likely.
  LIKELY = 4;

  // Very likely.
  VERY_LIKELY = 5;
}