AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

ClimbQ: Class Imbalanced Quantization Enabling Robustness on Efficient Inferences

Ting-An Chen, De-Nian Yang, Ming-syan Chen

NeurIPS 2022 · Label-efficient and robust learning › Recognition and long-tail · approach: Group-robust classification · 3 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add HomoVar loss on top of ClimbQ (ablation)ClimbQ · CIFAR-10-LT 2-bit MobileNet-V2, gamma=10accuracy (%): 69.94 -> 72.51Improved
add HomoVar loss on top of ClimbQ (ablation)ClimbQ · CIFAR-10-LT 2-bit MobileNet-V2, gamma=200accuracy (%): 52.47 -> 56.89Improved
add HomoVar loss on top of ClimbQ (ablation)ClimbQ · CIFAR-100-LT 2-bit ResNet-56, gamma=200accuracy (%): 23.91 -> 27.39Improved
add HomoVar loss on top of ClimbQ (ablation)ClimbQ · CIFAR-10-LT 4-bit ResNet-20, gamma=10accuracy (%): 80.72 -> 80.58No clear effect
replace BatchQuant with ClimbQ2-bit MobileNet-V2 · CIFAR-10-LT, gamma=10accuracy (%): 16.24 -> 69.94Improved
replace BatchQuant with ClimbQ+2-bit ResNet-56 · CIFAR-100-LT, gamma=10accuracy (%): 37.24 -> 49.12Improved
replace BatchQuant with ClimbQ+2-bit MobileNet-V2 · CIFAR-100-LT, gamma=200accuracy (%): 10.69 -> 25.45Improved
replace Choi et al. with ClimbQ (no HomoVar)4-bit MobileNet-V2 · CIFAR-100-LT, gamma=200accuracy (%): 28.30 -> 22.58Hurt
replace Choi et al. with ClimbQ (no HomoVar)4-bit MobileNet-V2 · CIFAR-10-LT, gamma=10accuracy (%): 80.13 -> 78.91Hurt
replace best baseline with ClimbQ+2-bit ResNet-20 · Syndigits-LT, gamma=200accuracy (%): 58.75 -> 80.55Improved
replace Focal loss rebalancing with HomoVar loss (ablation)ClimbQ 2-bit ResNet-20 · CIFAR-10-LT, gamma=200accuracy (%): 37.69 -> 61.06Improved
replace Class-Balanced (CB) rebalancing with HomoVar loss (ablation)ClimbQ 2-bit ResNet-20 · CIFAR-10-LT, gamma=50accuracy (%): 59.94 -> 71.73Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy