CasDyF-Net: Image Dehazing via Cascaded Dynamic Filters

He Bin; Wang Yinglong

RECORD

CasDyF-Net: Image Dehazing via Cascaded Dynamic Filters

Wang Yinglong, He Bin

Computer Vision and Pattern Recognition · Submitted 13 Sep 2024

abstract compiled · 2026-09-23 · no language-model review

The authors' main claim Experiments on RESIDE, Haze4K, and O-Haze datasets validate our method's effectiveness, with our model achieving a PSNR of 43.21dB on the RESIDE-Indoor dataset.

how it is stated · asserted 1.00 read the record →

Comment (0) Read on arXiv ↗

Cite ▾

Claim paper

The compiled record

The paper's own sentences, arranged. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. Numbers in the tables are the authors', checked by code. No sentence in this record was written by a model.

Open the compiled record

Compiled by classifier-1.13.0 via TypeSafe API, question bank abstract-p1-20260923, 2026-09-23: 1 request, 6,201 input tokens.

What the abstract itself states, as probabilities: claims first 0.05 · states limitation 0.08 · human participants 0.02 · controlled comparison 0.11 · randomized assignment 0.01 · states quantitative result 0.99 · compares with prior results 0.20 · mentions code or data availability 0.98.

The abstract, sentence by sentence

arXiv listing abstract v1. Results and the main claim are underlined; show every role.

Image dehazing aims to restore image clarity and visual quality by reducing atmospheric scattering and absorption effects. While deep learning has made significant strides in this area, more and more methods are constrained by network depth. Consequently, lots of approaches have adopted parallel branching strategies. however, they often prioritize aspects such as resolution, receptive field, or frequency domain segmentation without dynamically partitioning branches based on the distribution of input features. Inspired by dynamic filtering, we propose using cascaded dynamic filters to create a multi-branch network by dynamically generating filter kernels based on feature map distribution. To better handle branch features, we propose a residual multiscale block (RMB), combining different receptive fields. Furthermore, we also introduce a dynamic convolution-based local fusion method to merge features from adjacent branches. Experiments on RESIDE, Haze4K, and O-Haze datasets validate our method's effectiveness, with our model achieving a PSNR of 43.21dB on the RESIDE-Indoor dataset. The code is available at https://github.com/dauing/CasDyF-Net.

background contribution method result interpretation main claim

What the classifier reports for each sentence. The number is the probability it gave the label.
#	sentence	role	the authors' own work	states the main conclusion	strength
s0	Image dehazing aims to restore image clarity and visual quality by reducing atmospheric scattering and absorption effects.	background 1.00	0.03	0.03	asserted 1.00
s1	While deep learning has made significant strides in this area, more and more methods are constrained by network depth.	background 0.53	0.05	0.02	asserted 1.00
s2	Consequently, lots of approaches have adopted parallel branching strategies. however, they often prioritize aspects such as resolution, receptive field, or frequency domain segmentation without dynamically partitioning branches based on the distribution of input features.	gap 0.83	0.06	0.04	asserted 0.95
s3	Inspired by dynamic filtering, we propose using cascaded dynamic filters to create a multi-branch network by dynamically generating filter kernels based on feature map distribution.	contribution 0.80	0.84	0.10	asserted 0.95
s4	To better handle branch features, we propose a residual multiscale block (RMB), combining different receptive fields.	method 0.54	0.91	0.03	asserted 0.98
s5	Furthermore, we also introduce a dynamic convolution-based local fusion method to merge features from adjacent branches.	contribution 0.55	0.92	0.03	asserted 0.99
s6	Experiments on RESIDE, Haze4K, and O-Haze datasets validate our method's effectiveness, with our model achieving a PSNR of 43.21dB on the RESIDE-Indoor dataset.	result 1.00	0.97	0.61	asserted 1.00
s7	The code is available at https://github.com/dauing/CasDyF-Net.	contribution 0.53	0.48	0.02	asserted 0.97

What kind of paper the classifier reports this is: method or tool 1.00 · evidence from observation or experiment 0.97.

What was examined, and what was not

sources read

arXiv listing abstract v1

not examined

full text
figures and tables
references and citing sentences
relations to other papers

levels run

P abstract

not run

S sections
T body sentences
C citations
R cross-paper relations

Every quoted sentence on this record is the authors'. Every label is from Pith's closed vocabulary, with the probability shown beside it. Every number in the measurements table is the authors', checked by code. Record as JSON

The paper

Original abstract

Image dehazing aims to restore image clarity and visual quality by reducing atmospheric scattering and absorption effects. While deep learning has made significant strides in this area, more and more methods are constrained by network depth. Consequently, lots of approaches have adopted parallel branching strategies. however, they often prioritize aspects such as resolution, receptive field, or frequency domain segmentation without dynamically partitioning branches based on the distribution of input features. Inspired by dynamic filtering, we propose using cascaded dynamic filters to create a multi-branch network by dynamically generating filter kernels based on feature map distribution. To better handle branch features, we propose a residual multiscale block (RMB), combining different receptive fields. Furthermore, we also introduce a dynamic convolution-based local fusion method to merge features from adjacent branches. Experiments on RESIDE, Haze4K, and O-Haze datasets validate our method's effectiveness, with our model achieving a PSNR of 43.21dB on the RESIDE-Indoor dataset. The code is available at https://github.com/dauing/CasDyF-Net.

Human review

How Pith checks papers → · Full audit of this paper →

How the literature uses it

2 arXiv papers cite this in 2 sentences

Each citing sentence is taken from the citing paper's LaTeX source and typed by Pith's classifier. A label appears at probability 0.8 or more (0.95 for support), a contest at 0.7 or more; the quoted words are the citing authors'. How each label was checked · JSON

Papers citing it

Uses its method or tool 1 paper

In the second stage, we employed the image dehazing methods CasDyF-Net [ref] and SCANet [ref] .

4th Workshop on Maritime Computer Vision (MaCVi): Challenge Overview · Team NJUST-KMG · 1.00

Cites it as background 1 paper

Deep learning approaches have revolutionised this field with advances in deblurring [ref] , denoising [ref] , dehazing [ref] , and super-resolution [ref] tasks, enabling reliable recovery of image features across diverse degradation conditions.

Physics-Informed Graph Neural Networks for Frequency-Aware Optical Aberration Correction · Image Restoration Networks. · 0.96

What it cites · 9 sentences without a confident label

Uses the method or tool of 2 papers

We adopt the Enhanced Parallel Attention [ref] , proven feasible in literature, combined with multiple 1×1 convolutions as our Global Attention Fusion module.

MixDehazeNet : Mix Structure Block For Image Dehazing Network · Progressive Attention Fusion · 0.85
Finally, to visually observe the influence of the dilated convolution on the RMB receptive field, we use the effective receptive field (ERF) theory proposed in [ref] for visualization.

Understanding the Effective Receptive Field in Deep Convolutional Neural Networks · Qualitative analysis of RMB · 0.98

Compares with 1 paper

Compared to the novel dual-scale method LKD [ref] , our RMB also achieved a significant performance improvement, indicating the effectiveness of the RMB design.

LKD-Net: Large Kernel Convolution Network for Single Image Dehazing · Ablation Studies · 1.00

Cites as background 1 paper

Some models, like FFA-Net [ref] and MixDehazeNet [ref] , combine different attention types for better feature integration.

MixDehazeNet : Mix Structure Block For Image Dehazing Network · Attention Mechanisms · 0.89

Discussion

Know something about this paper? Leave a comment, or highlight any passage above to comment on it.