3,135 articles
DenseNet
DenseNet, short for Dense Convolutional Network, is a convolutional neural network architecture in which each layer connects to every preceding layer, introduced in 2016 and recognized with a CVPR 2017 Best Paper Award.
Density peak clustering
Density peak clustering is a clustering algorithm, proposed by Rodriguez and Laio in Science in 2014, that finds cluster centers as points of high density far from denser points.
Dependency grammar
Dependency grammar is an approach to syntax in which sentence structure consists of binary relations between a head and a dependent, founded by Lucien Tesnière's 1959 work.
Depth completion
Depth completion is the computer vision task of generating a dense pixel-wise depth map from sparse LiDAR or RGB-D measurements, used in autonomous driving, robotics, and augmented reality.
Depth-first search
Depth-first search (DFS), also called depth-first traversal, is a graph traversal algorithm that explores one path deeply before backtracking, and underlies cycle detection and topological sorting.
Derandomization
Derandomization converts a randomized algorithm into an equivalent deterministic one without substantially increasing its complexity; whether BPP equals P is a central open question in complexity theory.
Dereverberation
Dereverberation is a signal processing method that removes reverberation, the delayed reflections of sound in a room, from recordings to improve speech clarity and recognition.
Description logic
Description logic (DL) is a family of decidable knowledge representation languages describing domains via concepts, roles, and individuals, and underpinning the Web Ontology Language OWL 2.
Designing Data-Intensive Applications
Designing Data-Intensive Applications (DDIA) is Martin Kleppmann's 2017 O'Reilly book on database internals, replication, partitioning, transactions, consensus, and batch and stream processing; a second edition with Chris Riccomini appeared in February 2026.
Detached eddy simulation
Detached eddy simulation (DES) is a hybrid CFD approach that treats boundary layers with RANS equations and resolves detached eddies with LES, first proposed in 1997.
Detection Algorithms
Detection algorithms for object detection locate and classify objects in images via bounding boxes. Families: two-stage R-CNN series, one-stage YOLO/SSD, transformer-based DETR/RT-DETR; measured by mAP and FPS.
Detectron2
Detectron2 is an open-source Python library from Facebook AI Research, built on PyTorch, for object detection, instance segmentation, and other visual recognition tasks, and the successor of the original Detectron.
Determining the number of clusters in a data set
Determining the number of clusters in a data set is the problem of choosing how many groups a cluster analysis should produce, guided by indices like the silhouette.
Deterministic algorithm
A deterministic algorithm is an algorithm that, given the same inputs, always produces the same outputs on every run, making its behavior completely predictable from the input alone.
Deterministic simulation
A deterministic simulation is a computational model in which every input is fixed and non-random, so repeated runs with the same inputs give identical outputs.
DETR
DETR (DEtection TRansformer) is an object detection model introduced in May 2020 by Facebook AI researchers, treating detection as direct set prediction via a transformer encoder-decoder trained end-to-end.
DexYCB
DexYCB is a computer-vision dataset of human hands grasping real objects, created by NVIDIA and the University of Washington and presented at CVPR in 2021.
Dialogue state tracking
Dialogue state tracking (DST) is the task in task-oriented dialogue systems of estimating the user's goal each turn as (slot, value) pairs that guide the system's next action.
Dictionary learning
Dictionary learning is a machine learning method that learns basis elements, called atoms, so each signal can be represented sparsely as a combination of a few atoms.
Dictionary.com
Dictionary.com is a free, advertising-supported online English dictionary launched in 1995 by Brian Kariger and Daniel Fierro, based on the Random House Unabridged Dictionary and serving over 40 million users monthly.
DiffDock
DiffDock is a diffusion generative model developed at MIT that predicts the binding pose of a small-molecule ligand to a protein, a task known as molecular docking.
Differentiable architecture search
Differentiable architecture search (DARTS) is a neural architecture search method, described in 2018, that relaxes discrete architecture choices into a continuous form optimizable by gradient descent, cutting search cost to a few GPU days.
Differentiable optimization
Differentiable optimization embeds an optimization problem as a neural network layer: the forward pass solves it, and implicit differentiation of the optimality conditions supplies gradients for end-to-end training.
Differentiable programming
Differentiable programming is a programming paradigm in which programs are written so automatic differentiation, which applies the chain rule, computes exact gradients, allowing whole programs to be optimized end to end.
Differentiable simulation
Differentiable simulation is an approach to physics simulation in which the simulator is built from differentiable operations, supplying gradients for system identification, controller design, and policy learning.
Differential coding
Differential coding is a data compression technique that stores values as differences from a reference or previous value, so correlated data becomes smaller and easier to compress.
Differential evolution
Differential evolution (DE) is a population-based, gradient-free optimization algorithm introduced by Rainer Storn and Kenneth Price in 1995 and published in 1997, widely used for nonlinear and non-differentiable functions.
Differential privacy
Differential privacy is a mathematical definition of privacy for data analysis, formalized in 2006, guaranteeing that results barely change whether or not any one person's data is included.
Differential privacy SGD
Differentially private stochastic gradient descent (DP-SGD) is a privacy-preserving variant of stochastic gradient descent, introduced by Abadi et al. in 2016 and the standard method for private neural network training.
Differentially private learning
Differentially private learning trains machine learning models while mathematically limiting how much any single individual's data can influence the model, by adding calibrated noise during optimization.