Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2401.13174 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911100930359296 |
|---|---|
| author | Zhang, Dong Dong, Pingcheng Chen, Long Cheng, Kwang-Ting |
| author_facet | Zhang, Dong Dong, Pingcheng Chen, Long Cheng, Kwang-Ting |
| contents | It has been revealed that efficient dense image prediction (EDIP) models designed for AI chips, trained using the knowledge distillation (KD) framework, encounter two key challenges, including \emph{maintaining boundary region completeness} and \emph{ensuring target region connectivity}, despite their favorable real-time capacity to recognize the main object regions. In this work, we propose a customized boundary and context knowledge distillation (BCKD) method for EDIPs, which facilitates the targeted KD from large accurate teacher models to compact small student models. Specifically, the \emph{boundary distillation} focuses on extracting explicit object-level boundaries from the hierarchical feature maps to enhance the student model's mask quality in boundary regions. Meanwhile, the \emph{context distillation} leverages self-relations as a bridge to transfer implicit pixel-level contexts from the teacher model to the student model, ensuring strong connectivity in target regions. Our proposed method is specifically designed for the EDIP tasks and is characterized by its simplicity and efficiency. Theoretical analysis and extensive experimental results across semantic segmentation, object detection, and instance segmentation on five representative datasets demonstrate the effectiveness of BCKD, resulting in well-defined object boundaries and smooth connecting regions. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2401_13174 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Towards Customized Knowledge Distillation for Chip-Level Dense Image Predictions Zhang, Dong Dong, Pingcheng Chen, Long Cheng, Kwang-Ting Computer Vision and Pattern Recognition F.2.2 It has been revealed that efficient dense image prediction (EDIP) models designed for AI chips, trained using the knowledge distillation (KD) framework, encounter two key challenges, including \emph{maintaining boundary region completeness} and \emph{ensuring target region connectivity}, despite their favorable real-time capacity to recognize the main object regions. In this work, we propose a customized boundary and context knowledge distillation (BCKD) method for EDIPs, which facilitates the targeted KD from large accurate teacher models to compact small student models. Specifically, the \emph{boundary distillation} focuses on extracting explicit object-level boundaries from the hierarchical feature maps to enhance the student model's mask quality in boundary regions. Meanwhile, the \emph{context distillation} leverages self-relations as a bridge to transfer implicit pixel-level contexts from the teacher model to the student model, ensuring strong connectivity in target regions. Our proposed method is specifically designed for the EDIP tasks and is characterized by its simplicity and efficiency. Theoretical analysis and extensive experimental results across semantic segmentation, object detection, and instance segmentation on five representative datasets demonstrate the effectiveness of BCKD, resulting in well-defined object boundaries and smooth connecting regions. |
| title | Towards Customized Knowledge Distillation for Chip-Level Dense Image Predictions |
| topic | Computer Vision and Pattern Recognition F.2.2 |
| url | https://arxiv.org/abs/2401.13174 |