An Empirical Study of Adder Neural Networks for Object Detection

Authors: Xinghao Chen, Chang Xu, Minjing Dong, Chunjing XU, Yunhe Wang

NeurIPS 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental In this paper, we present an empirical study of Adder Nets for object detection. We present extensive ablation studies to explore several design choices of adder detectors. Comparisons with state-of-the-arts are conducted on COCO and PASCAL VOC benchmarks.
Researcher Affiliation Collaboration 1Huawei Noah s Ark Lab 2School of Computer Science, University of Sydney
Pseudocode No No pseudocode or clearly labeled algorithm blocks were found in the paper.
Open Source Code No The paper does not provide any statement or link regarding the public release of source code for the described methodology.
Open Datasets Yes We conduct experiments on the bounding box detection track of MS COCO 2017 and PASCAL VOC benchmarks, which have 80 and 20 object classes, respectively.
Dataset Splits Yes for COCO benchmark we use the COCO train2017 split that contains 118k images for training, val2017 split for validation (5k) and test-dev split (20k) for testing.
Hardware Specification No The paper mentions using GPUs (e.g., 'on one GPU', 'over 8 GPUs') but does not provide specific models or other hardware details for the experimental setup.
Software Dependencies No The paper mentions 'MMDetection [6]' as the base framework for implementation, but does not specify its version number or any other software dependencies with version numbers.
Experiment Setup Yes All models are trained with stochastic gradient descent (SGD) over 8 GPUs. Unless otherwise specified, all models are trained for 12 epochs (also known as 1 schedule) with cosine learning rate decay strategy. Weight decay and momentum are set to 0.0001 and 0.9, respectively.