DCIC uses dual constraints on a diffusion decoder to realize adjustable RDP operating points in neural image compression without extra rate cost.
hub
Variational image compression with a scale hyperprior
23 Pith papers cite this work. Polarity classification is still indexing.
23
Pith papers citing it
abstract
We describe an end-to-end trainable model for image compression based on variational autoencoders. The model incorporates a hyperprior to effectively capture spatial dependencies in the latent representation. This hyperprior relates to side information, a concept universal to virtually all modern image codecs, but largely unexplored in image compression using artificial neural networks (ANNs). Unlike existing autoencoder compression methods, our model trains a complex prior jointly with the underlying autoencoder. We demonstrate that this model leads to state-of-the-art image compression when measuring visual quality using the popular MS-SSIM index, and yields rate-distortion performance surpassing published ANN-based methods when evaluated using a more traditional metric based on squared error (PSNR). Furthermore, we provide a qualitative comparison of models trained for different distortion metrics.
hub tools
citation-role summary
background 3
method 1
citation-polarity summary
representative citing papers
GVCC achieves the lowest LPIPS on UVG at bitrates down to 0.003 bpp by encoding stochastic innovations in a marginal-preserving stochastic process derived from a pretrained rectified-flow video model, with 65% LPIPS reduction over DCVC-RT.
DeepCMC is a convolutional autoencoder architecture that compresses CSI matrices while jointly optimizing compression rate and reconstruction quality, outperforming prior schemes at equivalent bit rates.
GS-SCNet unifies 3D Gaussian Splatting with a disparity-guided semantic codec and direct Gaussian parameter prediction for efficient real-time 3D video communications with strong generalization.
MambaRaw uses SSM-based context modeling with TileMambaBlock and EAR modules for efficient JPEG-guided 4K raw reconstruction, reporting 1.2-1.4 dB PSNR gains and 9% lower latency over baselines on Sony, Olympus, and Samsung datasets.
MoECodec replaces FFN layers with token-wise MoE plus stable routing and GShMLP experts to support multiple downstream tasks in a single image compression model.
ECC integrates hyperprior side information, channel-wise context, latent residual prediction, temporal modeling, and entropy skip into a learned entropy model, yielding 39.9% and 76.3% average BD-rate reductions on ViSQOL and PESQ over baselines.
Few-step generative models can be reformulated as lossy codecs in the reverse channel coding framework without retraining, yielding faster encoding/decoding on low-resolution image benchmarks.
Diffusion model trained on synthetic geological data and conditioned via a dedicated encoder generalizes from stylized normal faults to strike-slip, flower, and thrust nappe structures.
Develops a local tangent-space rate-distortion theory and eigenspace-overlap diagnostic showing when physics-aligned compression necessarily degrades standard fidelity due to misaligned sensitivity directions.
MCWC aligns permutation-symmetric blocks across layers to enable sequential prediction and residual entropy coding, improving rate-accuracy tradeoffs versus quantization and prior codecs on language and vision models.
Introduces a benchmark for VLMs on compressed images and a universal adaptor to improve performance across codecs and bitrates.
RDJSCC introduces CVIE and CCF mechanisms to leverage source correlation for better image reconstruction in distributed transmission over severe fading channels with imperfect CSI.
W-IR is the first watermarking framework to combine certified robustness via randomized smoothing in pixel and coordinate spaces with identity leakage mitigation via residual information loss minimization.
RDVQ enables joint rate-distortion optimization for vector-quantized generative image compression via differentiable codebook distribution relaxation and an autoregressive entropy model.
FrequencyFormer co-designs a multi-scale DCT tokenizer, LUT-based near-sensor hardware, and modified MIPI communication to enable frequency-domain ViT inference with up to 128x data reduction and 230x lower communication energy.
Proposes T-GEMs plus a rate-based regularizer for early exits in CLIP encoders guided by text semantics to lower encoder usage costs.
ML-CLIPSim aggregates multi-layer patch and global similarities from frozen CLIP to approximate machine utility for images and outperforms standard IQA metrics on machine-preference tasks while staying competitive on human data.
A generative compression model using historical priors for Earth observation data achieves up to 10,000x reduction after exascale training on an Armv9 supercomputer.
SAMIC introduces semantic-aware Mamba blocks and SVD-based redundancy reduction to achieve efficient perceptual image compression with improved rate-distortion-perception tradeoffs.
PAT-VCM adds lightweight auxiliary tokens to a shared baseline video stream to support multiple downstream machine tasks without task-specific codecs.
Optimized JPEG AI (plus AVIF/WebP), grayscale or color choice, downscaling and mild blur keep face recognition viable at ≤1024-byte images for ICAO-style and ABC-gate scenarios.
Autoencoder CSI compression reduces channel sounding overhead by more than 50% versus standard IEEE 802.11 methods and improves throughput and latency in coordinated beamforming.
citing papers explorer
-
Dual-Constrained Diffusion Image Compression for Operational Rate-Distortion-Perception Optimization
DCIC uses dual constraints on a diffusion decoder to realize adjustable RDP operating points in neural image compression without extra rate cost.
-
GVCC: Zero-Shot Video Compression via Codebook-Driven Stochastic Rectified Flow
GVCC achieves the lowest LPIPS on UVG at bitrates down to 0.003 bpp by encoding stochastic innovations in a marginal-preserving stochastic process derived from a pretrained rectified-flow video model, with 65% LPIPS reduction over DCVC-RT.
-
Deep Convolutional Compression for Massive MIMO CSI Feedback
DeepCMC is a convolutional autoencoder architecture that compresses CSI matrices while jointly optimizing compression rate and reconstruction quality, outperforming prior schemes at equivalent bit rates.
-
Generalizable 3D Gaussian Splatting enabled Semantic Coding for Real-Time Immersive Video Communications
GS-SCNet unifies 3D Gaussian Splatting with a disparity-guided semantic codec and direct Gaussian parameter prediction for efficient real-time 3D video communications with strong generalization.
-
MambaRaw: Selective State Space Modeling for Efficient 4K Raw Image Reconstruction
MambaRaw uses SSM-based context modeling with TileMambaBlock and EAR modules for efficient JPEG-guided 4K raw reconstruction, reporting 1.2-1.4 dB PSNR gains and 9% lower latency over baselines on Sony, Olympus, and Samsung datasets.
-
MoECodec: Image Compression for joint human and machine perception via Mixture-of-Experts
MoECodec replaces FFN layers with token-wise MoE plus stable routing and GShMLP experts to support multiple downstream tasks in a single image compression model.
-
Benchmarking Neural Speech Compression from a Rate-Distortion Perspective
ECC integrates hyperprior side information, channel-wise context, latent residual prediction, temporal modeling, and entropy skip into a learned entropy model, yielding 39.9% and 76.3% average BD-rate reductions on ViSQOL and PESQ over baselines.
-
Few-step Generative Models as Lossy Compression
Few-step generative models can be reformulated as lossy codecs in the reverse channel coding framework without retraining, yielding faster encoding/decoding on low-resolution image benchmarks.
-
Implicit Structural Modeling via Generative Diffusion Frameworks
Diffusion model trained on synthetic geological data and conditioned via a dedicated encoder generalizes from stylized normal faults to strike-slip, flower, and thrust nappe structures.
-
A Geometric Lens on Physics-Aligned Data Compression
Develops a local tangent-space rate-distortion theory and eigenspace-overlap diagnostic showing when physics-aligned compression necessarily degrades standard fidelity due to misaligned sensitivity directions.
-
Motion-Compensated Weight Compression
MCWC aligns permutation-symmetric blocks across layers to enable sequential prediction and residual entropy coding, improving rate-accuracy tradeoffs versus quantization and prior codecs on language and vision models.
-
Benchmarking and Enhancing VLM for Compressed Image Understanding
Introduces a benchmark for VLMs on compressed images and a universal adaptor to improve performance across codecs and bitrates.
-
Robust Deep Joint Source-Channel Coding Enabled Distributed Image Transmission with Imperfect Channel State Information
RDJSCC introduces CVIE and CCF mechanisms to leverage source correlation for better image reconstruction in distributed transmission over severe fading channels with imperfect CSI.
-
"Training robust watermarking model may hurt authentication!'' Exploring and Mitigating the Identity Leakage in Robust Watermarking
W-IR is the first watermarking framework to combine certified robustness via randomized smoothing in pixel and coordinate spaces with identity leakage mitigation via residual information loss minimization.
-
Differentiable Vector Quantization for Rate-Distortion Optimization of Generative Image Compression
RDVQ enables joint rate-distortion optimization for vector-quantized generative image compression via differentiable codebook distribution relaxation and an autoregressive entropy model.
-
FrequencyFormer: A Co-Designed Sensor-to-Processor Pipeline for Frequency-Domain Vision Transformer Inference
FrequencyFormer co-designs a multi-scale DCT tokenizer, LUT-based near-sensor hardware, and modified MIPI communication to enable frequency-domain ViT inference with up to 128x data reduction and 230x lower communication energy.
-
t-gems: text-guided exit modules for decreasing clip image encoder
Proposes T-GEMs plus a rate-based regularizer for early exits in CLIP encoders guided by text semantics to lower encoder usage costs.
-
ML-CLIPSim: Multi-Layer CLIP Similarity for Machine-Oriented Image Quality
ML-CLIPSim aggregates multi-layer patch and global similarities from frozen CLIP to approximate machine utility for images and outperforms standard IQA metrics on machine-preference tasks while staying competitive on human data.
-
Transforming the Use of Earth Observation Data: Exascale Training of a Generative Compression Model with Historical Priors for up to 10,000x Data Reduction
A generative compression model using historical priors for Earth observation data achieves up to 10,000x reduction after exascale training on an Armv9 supercomputer.
-
SAMIC: A Lightweight Semantic-Aware Mamba for Efficient Perceptual Image Compression
SAMIC introduces semantic-aware Mamba blocks and SVD-based redundancy reduction to achieve efficient perceptual image compression with improved rate-distortion-perception tradeoffs.
-
PAT-VCM: Plug-and-Play Auxiliary Tokens for Video Coding for Machines
PAT-VCM adds lightweight auxiliary tokens to a shared baseline video stream to support multiple downstream machine tasks without task-specific codecs.
-
Optimizing Image Preparation and Compression for Face Recognition within 1024 Bytes
Optimized JPEG AI (plus AVIF/WebP), grayscale or color choice, downscaling and mild blur keep face recognition viable at ≤1024-byte images for ICAO-style and ABC-gate scenarios.
-
Autoencoder-Based CSI Compression for Beyond Wi-Fi 8 Coordinated Beamforming
Autoencoder CSI compression reduces channel sounding overhead by more than 50% versus standard IEEE 802.11 methods and improves throughput and latency in coordinated beamforming.