Research map / Architectures and efficiency / CNN design
Tensor Regression Networks
Jean Kossaifi, Zachary C. Lipton, Arinbjorn Kolbeinsson, Aran Khanna, Tommaso Furlanello, Anima Anandkumar
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 changed | Where | Measured | Effect |
|---|---|---|---|
| replace the fully connected layer with a tensor regression layer (256,3,4,3,1) | 3D-ResNet · UK Biobank MRI age | MAE (years): 2.96 -> 2.70 | Improved |
| replace the fully connected layer with a TRL for gender classification | 3D-ResNet · UK Biobank MRI gender | classification error (%): 0.79 -> 0.53 | Improved |
| replace the fully connected layer with a TRL for BMI regression | 3D-ResNet · UK Biobank MRI BMI | MAE (kg/m2): 2.37 -> 2.26 | Improved |
| use the full-rank 6x7x6 activation tensor in the TRL instead of low rank (ablation) | TRL on UK Biobank age · UK Biobank MRI age | MAE (years): 2.71 | No clear effect |
| remove average pooling without a TRL (keep spatial dims with FC) (ablation) | 3D-ResNet · UK Biobank MRI | training stability | Hurt |
| reduce TRL rank from (200,1,1,200) to (100,1,1,100) (ablation) | TRL on ResNet-101 · ImageNet-1k val | Top-1 accuracy (%): 77.1 -> 74.6 | Hurt |
| reduce TRL rank to (50,1,1,50) (ablation) | TRL on ResNet-101 · ImageNet-1k val | Top-1 accuracy (%): 77.1 -> 73.6 | Hurt |
| replace the FC layer with a TRL of rank (300,1,1,700) | ResNet-101 · ImageNet-1k val | Top-1 accuracy (%): 77.1 -> 77.2 | Improved |
| add a TCL (512,7,7) before the fully connected layers | VGG-19 · ImageNet-1k val | Top-1 accuracy (%): 68.7 -> 69.4 | Improved |
| shrink the TCL to (384,5,5) with 3072 hidden units | VGG-19 · ImageNet-1k val | Top-1 accuracy (%): 68.7 -> 68.3 | No clear effect |
| replace pooling and FC with an overcomplete TRL (1000,2048,7,7) | ResNet-50 · ImageNet-1k val | Top-1 accuracy (%): 74.58 -> 73.6 | Hurt |
| add a TCL (1024,3,3) before the TRL (ablation) | ResNet-50 with TRL (1000,1024,3,3) · ImageNet-1k val | Top-1 accuracy (%): 72.16 -> 73.43 | Mixed |
Closest papers by meaning
- Convolutions and More as Einsum: A Tensor Network Perspective with Advances for Second-Order Methods (NeurIPS 2024 · similarity 0.89)
- Geometry-aware training of factorized layers in tensor Tucker format (NeurIPS 2024 · similarity 0.87)
- Factor Augmented Tensor-on-Tensor Neural Networks (AAAI 2025 · similarity 0.87)
- Low-Rank Tensor Transitions (LoRT) for Transferable Tensor Regression (ICML 2025 · similarity 0.87)
- Distribution-Aware Tensor Decomposition for Compression of Convolutional Neural Networks (NeurIPS 2025 · similarity 0.87)
- SVDinsTN: A Tensor Network Paradigm for Efficient Structure Search from Regularized Modeling Perspective (CVPR 2024 · similarity 0.86)
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.
