Research map / Architectures and efficiency / RNNs and normalization
Energy-Constrained Compression for Deep Neural Networks via Weighted Sparse Projection and Layer Input Masking
Haichuan Yang, Yuhao Zhu, Ji Liu
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 |
|---|---|---|---|
| Use energy-constrained training with weighted sparse projection and input masking instead of MP | AlexNet · ImageNet | accuracy drop: 0.7% -> 0.5% | Improved |
| Use energy-constrained training with weighted sparse projection and input masking instead of MP | AlexNet · ImageNet | energy: 34% -> 26% | Improved |
| Use energy-constrained training with weighted sparse projection and input masking instead of EAP | SqueezeNet · ImageNet | energy: 76% -> 38% | Improved |
| Use energy-constrained training with weighted sparse projection and input masking instead of SSL | MobileNetV2 · ImageNet | accuracy drop: 2.0% -> 1.0% | Improved |
| Use energy-constrained training instead of NetAdapt | LeNet-5 · MNIST | accuracy drop: 0.6% -> 0.5% | Improved |
| Use energy-constrained training instead of SSL | MobileNetV2 · MS-Celeb-1M | accuracy drop: 0.7% -> 0.2% | Improved |
| Add knowledge distillation to MP (ablation) | MP on AlexNet · ImageNet | accuracy drop: 2.6% -> 0.7% | Improved |
| Add knowledge distillation to SSL (ablation) | SSL on AlexNet · ImageNet | accuracy drop: 17.6% -> 2.4% | Improved |
| Add knowledge distillation to MP (ablation) | MP on SqueezeNet · ImageNet | accuracy drop: 1.9% -> 1.7% | Improved |
| Add knowledge distillation to SSL (ablation) | SSL on SqueezeNet · ImageNet | accuracy drop: 16.0% -> 2.7% | Improved |
| Add knowledge distillation to MP (ablation) | MP on MobileNetV2 · ImageNet | accuracy drop: 2.0% -> 1.7% | Improved |
| Add knowledge distillation to SSL (ablation) | SSL on MobileNetV2 · ImageNet | accuracy drop: 1.9% -> 2.0% | No clear effect |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- NetAdapt — NetAdapt: Platform-Aware Neural Network Adaptation for Mobile Applications (ECCV 2018) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- ECC: Platform-Independent Energy-Constrained Deep Neural Network Compression via a Bilinear Regression Model (CVPR 2019 · similarity 0.91)
- Discrete Model Compression With Resource Constraint for Deep Neural Networks (CVPR 2020 · similarity 0.86)
- Designing Energy-Efficient Convolutional Neural Networks Using Energy-Aware Pruning (CVPR 2017 · similarity 0.86)
- "Lossless" Compression of Deep Neural Networks: A High-dimensional Neural Tangent Kernel Approach (NeurIPS 2022 · similarity 0.86)
- TinyTrain: Resource-Aware Task-Adaptive Sparse Training of DNNs at the Data-Scarce Edge (ICML 2024 · similarity 0.85)
- Towards training digitally-tied analog blocks via hybrid gradient computation (NeurIPS 2024 · similarity 0.84)
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.
