This calculator will help you estimate the cost of AI analysis of your media in our cloud. Use this calculator to prioritize media that justifies the AI spend.
Evolphin AI Hub allows you to specify rules that govern what media is recognized using AI. You can specify type of AI analysis to be performed on selected media based on metadata, folder or file name patterns.
Applies to: Video
Description
Returns Content Moderation label at a timecode in a video. Label categories include: Explicit Nudity, Suggestive, Violence, Visually Disturbing, Rude Gestures, Drugs, Tobacco, Alcohol, Gambling, Hate Symbols.
Example
“Name”: “Male Swimwear Or Underwear”,
“Category”: “Suggestive”,
“Time”: 0:30 seconds
Applies to: Audio, Video
Description
Profanity dictionary contains approximately 600 profane words and phrases. All the words transcribed from speech in audio or video that are included on this list will display will be displayed as asterisks except the first and last character. Editors can search in data timeline for all words that are filtered such as the ones with ** and then in an NLE like Premiere bleep out the audio using bars & tones effect easily.
Example
“F**k”,
“S**t”
Applies to: Audio, Video
Description
Convert speech in audio or video to text. Supported languages: English, Spanish, Portuguese, French and German. A data timeline with text phrases from audio speech is automatically created by Evolphin AI Hub. SRT & other caption formats can be used with downloads.
Example
“.. she said Evolphin’s AI prices are the best deal..”
Applies to: Video, Image
Description
Returns the bounding box for common object labels such as people, cars, furniture, apparel or pets in a video at a specific timecode or at a particular point in an image.
Example
“person”
“car”
“skateboard” at 1:45 mins or at x = 306px, y = 900px”
Applies to: Video, Image
Description
Returns the face bounding box as well as emotion, gender, age, pose, facial expressions. Note this engine does not attach name labels to faces. That is handled by separate engines such as Celebrity or Custom Face Recognition.
Example
“Female”, “Happy”, “26-43 years”, “Eyeglasses: Yes”, “Mustache: no”, “Mouth Open: Yes”, “Smile: Yes”
Applies to: Video, Image
Description
Returns name of detected celebrities in different settings, cosmetic makeup, and other conditions at a timecode in a video or at a point in an image. Also returns an IMDB URL for Celebrity Info that gets tagged in Evolphin Zoom metadata database.
Example
“Dua Lipa”
“Timestamp”: 49 seconds
“Location”: x= 200px, y = 90px”
Applies to: Video, Image
Description
Returns bounding boxes from detected text in video frames or images along with the actual text for indexing & searching in Zoom. The text or number can appear anywhere in the canvas or frame such as T-Shirts with text labels or jerseys with player number. To be detected, text must be within +/- 90 degrees orientation of the horizontal axis.
Example
“Golden Gate”
“Timestamp”: 01:21:20 hours
Applies to: Video
Description
Detect black frames, end credits, color bars or shot changes at a specific timecode in a video.
Example
“End Credits”” or “Shot Detected”
“Start Time: 1hr:28 min
“End Time””: 1hr:30 min”
Applies to: Video, Image
Description
Train the AI engine with your custom faces and then have Evolphin return the bounding boxes from custom faces detected in video frames or image canvas. Evolphin will index and allow you to then search your custom face name within a video. These can be any faces not just celebrities that you wish to recognize.
Example
“Jane Doe” , “My CEO”, “Dear Daughter”
“Timestamp”: 01:11:10 hours