Wiki

Frames per second (fps) is a measure of how many still images or frames are displayed in one second of a video or animation.

HITL is an iterative feedback process by which a person (or team) interacts with an algorithmically generated system (e.g., computer vision, machine learning, or artificial intelligence).

In machine learning, hyperparameters are given in advance to control the parameters of the learning process, while the values of other parameters (such as node weights) are obtained through training.

In terms of computer vision, diffusion models can be applied to a variety of tasks including image denoising, inpainting, super-resolution, and image generation.

In the field of deep learning, Ground Truth (commonly used in English, meaning "ground truth" or "benchmark truth" in Chinese, simply understood as the true value) refers to the accurate labels or data used to train and evaluate models.

Image Annotation is the process of tagging or annotating images with metadata, or additional information about the image content.

Human Pose Estimation (HPE) is a task in computer vision that involves detecting and estimating the positions of various body parts in images or videos of people.

An epoch in machine learning means the process of passing the entire training dataset through the neural network once (i.e., performing one forward propagation and one back propagation). For example, if the dataset consists of 1,000 samples and the model is trained using a batch size of 100, it will take 1 […]

False Positive Rate is a measure of the accuracy of a machine learning model in predicting positive outcomes. It is the proportion of instances where the model predicted a positive outcome but the actual outcome was negative.

A class boundary is the dividing line between two adjacent classes or categories in a dataset.

Concept drift refers to the phenomenon that the statistical properties of a data stream change over time, causing the learning model to not match the current data distribution.

Proximal Policy Optimization (PPO) is an algorithm in the field of reinforcement learning that is used to train the decision-making functions of computer agents to complete difficult tasks.

The Confusion Matrix is a performance evaluation tool used in machine learning that summarizes the performance of a classification model by listing the true positive, true negative, false positive, and false negative predictions.

Calibration curves are a useful tool in machine learning and predictive modeling to understand and fine-tune the reliability of a classification model's predicted probabilities.

Edge detection is a fundamental problem in image processing and computer vision. The purpose of edge detection is to identify points in digital images where brightness changes significantly.

In image processing and computer vision, the Laplacian operator has been used for various tasks such as blob detection and edge detection.

Differentiable Programming is a programming paradigm in which digital computer programs can be made fully differentiable via automatic differentiation.

Aspect-level sentiment analysis is a task to detect the sentiment of a specific aspect in a text.

Hallucination refers to the phenomenon that model-generated content is inconsistent with real-world facts or user input.

Foundation Agent is a general agent model that can be generalized in both the virtual world and the real world.

KV Cache is an important engineering technology for optimizing Transformer reasoning performance. This technology can improve reasoning performance by trading space for time without affecting any calculation accuracy.

Rotational Position Encoding (RoPE) is a position encoding method that can integrate relative position information dependency into self-attention and improve the performance of transformer architecture.

Virtual screening technology aims to search for potential drug molecules that interact with specific protein pockets from a large library of compounds through computational methods.

Floating-point operations per second (FLOPS) is a measure of computer performance based on the number of floating-point arithmetic calculations a processor can perform in one second.