Random Path Selection for Continual Learning

By Jathushan Rajasegaran, Munawar Hayat, Salman Hameed Khan, Fahad Shahbaz Khan, Ling Shao. 2019

168 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Meta & Continual Learning

Citations in the corpus

Continual Learning by Modeling Intra-Class Variation (2022, methodology): “, 2020), dynamic architecture-based methods (Rajasegaran et al., 2019; Hung et al., 2019; Yan et al., 2021) and memory-based methods (Rebuffi et al.”
Class Gradient Projection For Continual Learning (2022, background): “Prevalent approaches to address the catastrophic forgetting problem in continual learning can be grouped into three broad categories: regularization-based approaches [2, 22, 30, 45], expansion-based approaches [17, 25, 31, 34, 36, 40], and memory-based approaches [3, 27, 33, 38].”
Class-Incremental Learning with Cross-Space Clustering and Controlled Transfer (2022, neutral): “In dynamic architecture methods [55, 52, 31, 1, 41], the network architecture is modified by expansion or masking when encountering new tasks during learning.”
Online Continual Learning with Contrastive Vision Transformer (2022, neutral): “As new tasks arrive, expansion-based methods dynamically expand networks and keep sub-networks related to previous tasks fixed [69,89,1,59,86,45,82,53].”
DualPrompt: Complementary Prompting for Rehearsal-free Continual Learning (2022, methodology): “We compare against representative class-incremental learning methods, including DualNet [43], SupSup [61], DynaER [64] and RPSNet [48].”
General Incremental Learning with Domain-aware Categorical Representations (2022, methodology): “The third are the structure-based methods [24, 37], which allocate new parameters at each new session and prevent the change of the representation learned at previous sessions.”
Recent Advances of Continual Learning in Computer Vision: An Overview (2021, neutral): “Rajasegaran et al. [178] pointed out that the PathNet [146] requires heavy computational costs.”
An EM Framework for Online Incremental Learning of Semantic Segmentation (2021, neutral): “Structured-based methods[1, 27, 32] separates parameters learned at different steps from each other to avoid undesirable overlapping in representations.”
ACAE-REMIND for Online Continual Learning with Compressed Feature Replay (2021, methodology): “Also note that in this setting, we get competitive results compared with the methods BiC [35] and RPSnet [29].”
On Learning the Geodesic Path for Incremental Learning (2021, background): “[32] propose a simple controller with random search to choose the optimal paths for a new task.”
DER: Dynamically Expandable Representation for Class Incremental Learning (2021, methodology): “For class incremental learning, RPSNet [26] proposes a random path selection algorithm that progressively chooses optimal paths as sub-network for the new classes.”
Gradient Projection Memory for Continual Learning (2021, methodology): “In this setup, we believe a subset of old data replay either from storage or via generation is inevitable for attaining better per-8 formance with minimal forgetting (Rajasegaran et al., 2019).”
Training Networks in Null Space of Feature Covariance for Continual Learning (2021, positive): “An alternative approach is assigning different sub-networks to different tasks by masking the neurons [1, 41, 45] or weights [31].”
Online Continual Learning in Image Classification: An Empirical Survey (2021, methodology): “Most previous works require task-IDs at inference time, and a few recent methods have been introduced to predict without a taskID [19, 57].”
Do not Forget to Attend to Uncertainty while Mitigating Catastrophic Forgetting (2021, methodology): “In [31], the authors use the dropout based model selection for incremental learning.”
Adaptive Aggregation Networks for Class-Incremental Learning (2020, background): “[31] progressively chose the optimal paths for the new task while encouraging to share parameters across tasks.”
MgSvF: Multi-Grained Slow versus Fast Framework for Few-Shot Class-Incremental Learning (2020, background): “In a few-shot class-incremental learning setup, a network learns several tasks continually, and each task contains a batch of new classes [1], [9], [52].”