ApsaraVideo Media Processing:Parameters for intelligent production API operations

Name

Type

Description

Output

String

If the JobParams parameter is configured to separate Chinese and English, {resultType} placeholders are supported in the output file path to specify whether the output caption file is in Chinese or English. zh indicates Chinese and en indicates English.

Name

Type

Required

Description

fps

INT

No

The sampling frame rate. Valid values: [2,10]. Default value: 5.

roi

LIST

No

The region of interest.

• If specified, only text within the region is extracted. By default, text within the bottom quarter of the video frame is extracted.

• Set the value in the following format: [[top, bottom], [left, right]].

• Default value: N/A.

sep

BOOLEAN

No

Whether to generate separate Chinese and English SRT files. Default value: False.

formatter

String

No

The format string of the SRT caption. Example: "{\an8}". Default value: N/A.

track

String

No

If you set this parameter to "main", only the main caption track is extracted. If you do not specify this parameter, all captions that appear in the specified region are extracted by default. Default: None.

Name

Type

Description

Result

String

The detailed information about the job result. Example of success result information: {"Code":"Success","Message":"Successful.","Data":"{\"result\":[{\"file\":\"captionextraction/b48d02b58e9b6a0d1c13271bcf9aa6d7-161121379****.srt\"}]}"}.

Name

Type

Required

Description

bgimage

String

No

The replacement background image. Example: http://example-image-****.example-location.aliyuncs.com/example/example.jpg.

• If you specify this parameter, an MP4 video whose background image is replaced is returned.

• If you do not specify this parameter, a WebM video with alpha channels is returned.

Name

Type

Description

Result

String

The detailed information about the job result. Example of success result information: {"Code":"Success","Message":"Successful.","Data":"{\"result\":[{\"file\":\"videogreenscreenmatting/16e6bc5ca802e12429d082010164dba3-160275535****_matting.mp4\"}]}"}.

Name

Type

Required

Description

None

None

None

None

Name

Type

Description

Result

String

The detailed information about the job result. Example of success result information: {"Code":"Success","Data":"{\"result\":[{\"start\":39.32,\"end\":63.85,\"title\":\"Chorus\"},{\"start\":86.69,\"end\":114.45,\"title\":\"Chorus\"},{\"start\":135.75,\"end\":160.27,\"title\":\"Chorus\"}]}","Message":"Successful."}.

Name

Type

Required

Description

Text

LIST

No

The location of a caption box that you want to remove. A maximum of two caption boxes are supported. Example: [[bx1, by1, bw1, bh1], [bx2, by2, bw2, bh2]].

• bx: The ratio of the normalized x-coordinate of the upper-left corner of the caption box to the video width. Example: 0.1.

• by: The ratio of the normalized y-coordinate of the upper-left corner of the caption box to the video height. Example: 0.0.

• bw: The ratio of the normalized width of the caption box to the video width. Example: 0.3.

• bh: The ratio of the normalized height of the caption box to the video height. Example: 0.2.

LimitRegion

LIST

No

The area in which you want to remove captions. The system detects the captions within the specified area and removes the detected captions. This parameter has a lower priority than the Text parameter that directly specifies the location of caption boxes to be removed. Example: [[0, 0.6, 1, 0.4]]. In this example, the system detects the captions within the bottom 40% of the video image and removes the detected captions.

Name

Type

Description

Result

String

The detailed information about the job result. Example of success result information: {"Details":[],"Message":"success","Code":"Success"}.

Name

Type

Required

Description

None

None

None

None

Name

Type

Description

Result

STRING

The detailed information about the job result. Example of success result information: {"Details":[],"Message":"success","Code":"Success"}.

Name

Type

Required

Description

Logo

String

No

The position of a logo that you want to remove. Set the value in the format of [xmin, ymin, width, height]. You can remove up to two logos at a time. Example: [[0, 0, 0.3, 0.3], [0.7, 0, 0.3, 0.3]].

Name

Type

Description

Result

String

The detailed information about the job result. Example of success result information: {"Details":[],"Message":"success","Code":"Success"}.

Name

Type

Required

Description

Model

String

No

The smart thumbnail model. If left empty, a still thumbnail is generated. Set to gif to generate an animated thumbnail.

Name

Type

Description

Result

String

The detailed information about the job result. Example of success result information: {"Details":[],"Message":"success","Code":"Success"}{"Message":"success","Data":"[{\"Score\":8.270855992569906,\"Time\":\"28278.25\",\"Url\":\"cover/test-00001.jpg\"},{\"Score\":7.474117489692728,\"Time\":\"25942.583333333332\",\"Url\":\"cover/test-00002.jpg\"}]","Code":"Success"}. In this example, Score indicates the confidence of the thumbnail result, Time indicates the timestamp of the thumbnail frame, and Url indicates the URL of the thumbnail.

Name

Type

Required

Description

None

None

None

None

Name

Type

Description

Result

String

The detailed information about the job result. Example of success result information:

{"Code":"Success","Message":"Successful.","Data":"{\"result\":[{\"file\":\"videoclip/16e6bc5ca802e12429d082010164****-1602755353502-origin.mp4\"}]}"}

Name

Type

Required

Description

None

None

None

None

Name

Type

Description

Result

String

The details of the job result. This parameter is returned when the job is successful. Example: {"Details":[],"Message":"success","Code":"Success"}

Name

Type

Required

Description

None

None

None

None

Name

Type

Description

Result

STRING

The detailed information about the job result. Example of success result information: {"Details":[],"Message":"success","Code":"Success"}.

Name

Type

Required

Description

None

None

None

None

Name

Type

Description

Result

STRING

The detailed information about the job result. Example of success result information: {"Code":"Success","Data":"{\"result\":[{\"file\":\"detectresult/normalvideo-161225931****.txt\"}]}","Message":"Successful."}.

Name

Type

Required

Description

inputs

String

No

The URL of the audio track file to mix. Only one URL is supported. Example: {"file":"http://example-bucket-****.oss-cn-shanghai.aliyuncs.com/2.mp4"}.

Name

Type

Description

Result

String

The detailed information about the job result. Example of success result information: {"Message":"Successful.","Data":"{\"result\":[{\"file\":\"audiomix/alibaba-161283935****-origin.mp4\"}]}","Code":"Success"}.

Name

Type

Description

Output

STRING

{resultType} placeholders are supported in the path to distinguish whether the result file is a cartoonized image or the original image. result indicates a cartoonized image, and origin indicates the original image.

Name

Type

Description

Result

String

The detailed information about the job result. Example of success result information: {"Code":"Success","Data":"{\"result\":[{\"file\":\"iproduction/test-result.jpg\"},{\"file\":\"iproduction/test-origin.jpg\"}]}","Message":"Successful."}.

Name

Type

Description

Result

String

Details of a successful job result.

Parameter

Description

Time

The timestamp generated when the input file was scored.

Input

The input file name.

Total Duration

The duration of the input file. Unit: seconds.

Speech Ratio

The ratio of the duration of the audio data to the duration of the input file. Valid values: [0,100]. Unit: percentage.

Tag

A tag that indicates whether the detection is valid.

• Valid: The detection is valid, which indicates that subsequent key metrics and the mean opinion score (MOS) are valid.

• File too Short: The duration of the input file is less than 2s.

• Mute: The file contains only silence.

• Voice too Short: The duration of the audio data is less than 2s.

MOS(0-5)

The MOS of the input file, which describes the quality of the audio data. Valid values: [0,5].

MOS

The score level. Valid values:

• >4–5: Excellent

• [3, 4): Good

• [2, 3): Fair

• [1, 2): Poor

• [0, 1): Bad

Discontinuity(0-5)

The continuity score of the audio data. The continuity score decreases due to the following reasons: the stuttering issue of audio data capture, echo issue due to multi-channel audio, and packet loss issue due to poor network connectivity. Valid values: [0,5].

Discontinuity

The continuity score level. Valid values:

• (4, 5]: Excellent

• [3, 4): Good

• [2, 3): Fair

• [1, 2): Poor

• [0, 1): Poor

Coloration(0-5)

The intelligibility score of the audio data. The intelligibility score decreases due to the following reasons: large reverberation, low bitrate, encoding error, and ambiguous pronunciation. Valid values: [0,5].

Coloration

The score level. Valid values:

• (4, 5]: Excellent

• [3, 4): Good

• [2, 3): Fair

• [1, 2): Poor

• [0, 1): Bad

Noisiness(0-5)

The noise score of the audio data. Valid values: [0,5].

Noisiness

The score level. Valid values:

• (4, 5]: Excellent

• [3, 4): Good

• [2, 3): Fair

• [1, 2): Poor

• [0, 1): Poor

Loudness(0-5)

The loudness score of human voice. Higher values indicate clearer and louder voice; lower values indicate barely audible voice. Valid values: [0,5].

Loudness

The score level. Valid values:

• (4, 5]: Excellent

• [3, 4): Good

• [2, 3): Fair

• [1, 2): Poor

• [0, 1): Poor

Loudness(-90dB-0dB)

The average volume of the human voice. Valid values: [-90,0]. Unit: decibel.

• Describes voice volume in decibels. Values below -24 typically indicate low voice volume.

• Default value: -90.0, indicating that no human voice is detected.

Double Talk Ratio(%)

The double talk ratio helps determine the possible causes of a low Discontinuity score. The value is a percentage that ranges from 0 to 100.

Bad Mute Ratio(%)

The percentage of abnormal mute frames. All abnormal mute frames of the audio data that does not include two-channel audio data are counted, excluding mute frames caused by cutting two-channel audio data. Valid values: [0,100]. Unit: percentage.

Saturated Ratio(%)

This metric indicates the percentage of a speech segment that contains popping sounds. It helps diagnose widespread popping issues caused by an excessively high recording volume. The value is a percentage that ranges from 0 to 100.

Worst MOS(0-5)

The lowest MOS during the scoring process. Valid values: [0,5].

Worst Discontinuity(0-5)

The worst discontinuity score during the scoring process. Valid values: [0, 5].

Worst Noisiness(0-5)

The highest noise score during the scoring process. Valid values: [0,5].

Worst Coloration(0-5)

The worst coloration score during the scoring process. Valid values: [0, 5].

Name

Type

Required

Description

beauty_params

String

No

The parameters of the FaceBeauty operation. Example: "whiten=20,smooth=50,face_thin=50"

Parameter

Type

Description

skin_beauty_enable

int

Whether to enable skin polishing.

• Valid values: [0,1].

• 0: disabled

• 1: enabled

• Default value: 1.

shape_beauty_enable

int

Whether to enable face shape adjustment.

• Valid values: [0,1].

• 0: disabled

• 1: enabled

• Default value: 1.

whiten

INT

The degree of skin whitening. The greater the value, the whiter the skin looks.

• Valid values: [0,100].

• Default value: 20

smooth

int

The degree of skin smoothing. The greater the value, the more smooth the skin looks.

• Valid values: [0,100].

• Default value: 20

detail

INT

The degree of skin granularity. The greater the value, the more fine-grained the skin is, and the more skin details exist.

• Valid values: [0,100].

• Default value: 20

skin_model

int

Whether to enable the skin model feature. When enabled, skin whitening applies only to areas detected as skin.

• Valid values: [0,1].

• 0: disabled

• 1: enabled

• Default value: 1.

cheek_thin

float

The degree of frontal bone thinning.

• Valid values: [0,100].

• Default value: 0.

face_cut

float

The degree of cheekbone narrowing.

• Valid values: [0,100].

• Default value: 0.

face_thin

float

The degree of face thinning.

• Valid values: [0,100].

• Default value: 0.

face_length

float

The degree of face length adjustment (two-way).

• Valid values: [-100,100].

• Default value: 0.

chin_length

float

The degree of chin length adjustment (two-way).

• Valid values: [-100,100].

• Default value: 0.

chin_thin

float

The degree of chin thinning.

• Valid values: [0,100].

• Default value: 0.

eye_size

float

The degree of eye enlargement.

• Valid values: [0,100].

• Default value: 0.

eye_corner1

float

The degree of vertical canthus adjustment (two-way).

• Valid values: [-100,100].

• Default value: 0.

eye_distance

float

The degree of eye distance adjustment (two-way).

• Valid values: [-100,100].

• Default value: 0.

nose_thin

float

The degree of nose slimming (two-way).

• Valid values: [-100,100].

• Default value: 0.

nose_wing

float

The degree of nasal alar slimming (two-way).

• Valid values: [-100,100].

• Default value: 0.

nose_length

float

The degree of nose length adjustment (two-way).

• Valid values: [-100,100].

• Default value: 0.

mouth_size

FLOAT

The degree of mouth size adjustment (two-way).

• Valid values: [-100,100].

• Default value: 0.

mouth_position

float

The degree of philtrum length adjustment (two-way).

• Valid values: [-100,100].

• Default value: 0.

lip_thickness

float

The degree of lip thickness adjustment (two-way).

• Valid values: [-100,100].

• Default value: 0.

hair_line

float

The degree of hairline adjustment (two-way).

• Valid values: [-100,100].

• Default value: 0.

smile

float

The degree of smile enhancement.

• Valid values: [0,100].

• Default value: 0.

detect_mode

float

The face detection mode.

• Valid values: [0,1].

• 0: video mode.

• 1: image mode.

• Default value: 1.

detect_level

float

The resolution of the face detector. Smaller faces may not be detected at lower resolutions.

• Valid values: [0,2].

• 0: the lowest resolution at the fastest detection speed.

• 1: the medium resolution at medium detection speed.

• 2: the highest resolution at the slowest detection speed.

• Default value: 1.

threshold

float

The confidence threshold for face detection.

• Valid values: [0,1].

• Default value: 0.8.

detect_interval

float

The number of frames between consecutive face detections in video mode.

• Valid values: [1,65535].

• Default value: 5.

max_face_num

float

The maximum number of detectable faces.

• Valid values: [0,32].

• Default value: 32.

min_face

float

The minimum face width in pixels.

• Valid values: [10,1024].

• Default value: 40.

Name

Type

Description

Result

String

The detailed information about the job result. Example of success result information: {"Code":"Success","Data":"{\"result\":[{\"file\":\"result.mp4\"}]}","Message":"Successful."}.