Research map / Architectures and efficiency / Pruning and efficient tuning
BATUDE: Budget-Aware Neural Network Compression Based on Tucker Decomposition
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 changed | Where | Measured | Effect |
|---|---|---|---|
| Use budget-aware, globally learned Tucker ranks instead of fixed equal ranks | ResNet-20 · CIFAR-10 | Top-1 accuracy: 87.41 -> 91.02 | Helped |
| Use budget-aware Tucker ranks rather than genetic-algorithm tensor-rank selection | ResNet-20 · CIFAR-10 | Top-1 accuracy: 88.50 -> 89.47 | Helped |
| Use budget-aware Tucker ranks rather than genetic-algorithm tensor-rank selection | ResNet-32 · CIFAR-10 | Top-1 accuracy: 90.6 -> 91.47 | Helped |
| Use budget-aware Tucker ranks rather than genetic-algorithm tensor-rank selection | ResNet-20 · CIFAR-100 | Top-1 accuracy: 66.13 -> 66.67 | Helped |
| Use budget-aware Tucker ranks rather than genetic-algorithm tensor-rank selection | ResNet-32 · CIFAR-100 | Top-1 accuracy: 68.05 -> 68.95 | Helped |
| Use budget-aware Tucker ranks rather than fixed-rank Tucker decomposition | ResNet-18 · ImageNet | Top-5 accuracy: 87.53 -> 89.41 | Helped |
| Use budget-aware Tucker decomposition instead of the uncompressed model | ResNet-18 · ImageNet | Top-5 accuracy: 89.08 -> 89.41 | Helped |
| Use budget-aware Tucker decomposition instead of the uncompressed model | ResNet-50 · ImageNet | Top-5 accuracy: 92.87 -> 93.24 | Helped |
| Use budget-aware Tucker decomposition instead of the uncompressed model | ResNet-50 · ImageNet | Top-5 accuracy: 92.87 -> 93.42 | Helped |
Closest papers by meaning
- Towards Efficient Tensor Decomposition-Based DNN Model Compression With Optimization Framework (CVPR 2021 · similarity 0.91)
- Distribution-Aware Tensor Decomposition for Compression of Convolutional Neural Networks (NeurIPS 2025 · similarity 0.89)
- Compressing Neural Networks: Towards Determining the Optimal Layer-wise Decomposition (NeurIPS 2021 · similarity 0.88)
- Exploring Unexplored Tensor Network Decompositions for Convolutional Neural Networks (NeurIPS 2019 · similarity 0.87)
- Geometry-aware training of factorized layers in tensor Tucker format (NeurIPS 2024 · similarity 0.85)
- MCNC: Manifold-Constrained Reparameterization for Neural Compression (ICLR 2025 · similarity 0.85)
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.