Results for "anomaly detection"

Identifying suspicious transactions.

Detects trigger phrases in audio streams.

Identifying and localizing objects in images, often with confidence scores and bounding rectangles.

High-fidelity virtual model of a physical system.

Observing model inputs/outputs, latency, cost, and quality over time to catch regressions and drift.

Signals indicating dangerous behavior.

Hidden behavior activated by specific triggers, causing targeted mispredictions or undesired outputs.

Identifying abrupt changes in data generation.

AI focused on interpreting images/video: classification, detection, segmentation, tracking, and 3D understanding.

Pixel-level separation of individual object instances.

Of predicted positives, the fraction that are truly positive; sensitive to false positives.

Of true positives, the fraction correctly identified; sensitive to false negatives.

Of true negatives, the fraction correctly identified.

Networks using convolution operations with weight sharing and locality, effective for images and signals.

Automated detection/prevention of disallowed outputs (toxicity, self-harm, illegal instruction, etc.).

When some classes are rare, requiring reweighting, resampling, or specialized metrics.

Maliciously inserting or altering training data to implant backdoors or degrade performance.

Combining signals from multiple modalities.

Assigning labels per pixel (semantic) or per instance (instance segmentation) to map object boundaries.

Pixel motion estimation between frames.

Identifying speakers in audio.

Low-latency prediction per request.

Process for managing AI failures.

External sensing of surroundings (vision, audio, lidar).

Software simulating physical laws.

Detecting and avoiding obstacles.

Automated assistance identifying disease indicators.