Vision & Audio47
How AI recognizes objects in photos, tracks body poses and turns speech into text. CNNs, object detection, segmentation, pose estimation, TTS, STT: many of these you can try with just a webcam and a microphone.
Watching a run of frames to name what action is happening
Background RemovalErasing a photo's background and keeping just the subject
Bounding BoxThe smallest box around an object, and where that box sits
CNNConvolutional Neural NetworkA neural network built to spot patterns in a photo, layer by layer
ColorizationAdding fitting color to a black-and-white picture
Computer VisionTechnology that gets a computer to make sense of photos and video
Confidence ScoreThe number a model attaches to how sure it is about its own answer
ConvolutionSliding a small tile across a photo to score it, spot by spot
DenoisingRemoving only the flecks mixed in, leaving the original untouched
Depth EstimationWorking out near and far from a single flat photo
Face DetectionDrawing a box around every face in a photo
Face RecognitionComparing faces to work out who someone is
Feature MapA picture that lays scan results out in their original spots
Filter / KernelA small value tile that carries what to look for
Gaussian SplattingA 3D scene held as millions of scattered color blobs
Gesture RecognitionReading a hand or body movement as one predefined meaning
Hand TrackingFollowing hand joint locations from one moment to the next
Image CaptioningLooking at a picture and writing what's in it as a sentence
Image ClassificationPutting one label on an entire photo
KeypointA point that pins down one location in an image
LandmarkA numbered reference point at a fixed spot
Lip SyncFitting mouth shapes to the beat of a sound
Motion CaptureTurning body movement into a record of point locations
Music GenerationTurning a few lines of request into a finished piece of music
NeRFNeural Radiance FieldA 3D scene that answers with light for any spot and direction
Object DetectionFinding every object in a photo and boxing each one
Object TrackingKeeping the same number attached to the same thing over time
OCRReading the text in a picture and turning it into characters
Pattern RecognitionSpotting the regularity that keeps repeating
Pitch DetectionPinpointing how high or low a sound is
PoolingShrinking a picture by keeping one representative value per region
Pose EstimationMarking joint locations with points to work out a body's pose
Real-Time InferenceFinishing a judgment within a fixed window, every single moment
ResolutionThe number of dots a picture is made of
SegmentationPutting a label on every single pixel of a photo
Sound ClassificationSticking one label on a short clip of sound
Source SeparationSplitting sound that's been mixed together back into its parts
Speaker RecognitionTelling who spoke from the sound of their voice
SpectrogramA picture drawn so you can see sound with your eyes
Speech RecognitionThe technology that transcribes spoken words into text
STTSpeech-to-TextThe conversion that turns sound into written words
TTSText-to-SpeechThe technology that reads text out loud in a voice
UpscalingEnlarging a picture by inventing the detail it never had
Vision TransformerCutting a photo into pieces and treating it like text
VocoderThe part that turns a sound blueprint into an actual waveform
Voice AssistantAn AI you speak commands to, and hear answers from
Voice CloningImitating a person's voice from a short recording