MIEP: Channel Pruning with Multi-granular Importance Estimation for Object Detection

Jiaxin Chen

Abstract:

This paper investigates compressing a pre-trained deep object de
tector to a lightweight one by channel pruning, which has proved
effective and flexible in promoting efficiency. However, the majority
of existing workstrimchannelsbasedonamonotonouscriterionfor
general purposes, i.e., the importance to the task-specific loss. They
are prone to overly prune intermediate layers and simultaneously
leave large intra-layer redundancy, severely deteriorating the detec
tion accuracy. To address the issues above, we propose a novel chan
nel pruning approach with multi-granular importance estimation
(MIEP), consisting of the Feature-level Object-sensitive Importance
(FOI) and the Intra-layer Redundancy-aware Importance (IRI). The
former puts large weights on channels that are critical for object
representation through the guidance of object features from the pre
trained model, and mitigates over-pruning when combined with
the task-specific loss. The latter groups highly correlated channels
based on clustering, which are subsequently pruned with priority
to decrease redundancy. Extensive experiments on the COCO and
VOCbenchmarks demonstrate that MIEP remarkably outperforms
the state-of-the-art channel pruning approaches, achieves a better
balance between accuracy and efficiency compared to lightweight
object detectors, and generalizes well to various detection frame
works (e.g., Faster-RCNN and FSAF) and tasks (e.g., classification).