arXiv:2608.28730v1 Announce Type: cross Abstract: Latest JPEG restoration systems achieve strong quality with large models, yet often remain too slow and expensive for efficient on-device deployment....
The paper introduces Neural Image Compression with Internal-External Prior Fusion and Alignment (IE-PFA), a context model that combines internal priors from the current image with an external dictionary shared by encoder and decoder. It employs a RWKV-based Spatial-Channel Prior Fusion (RSCPF) module for efficient fusion and a Self-Distilled Prior Alignment (SDPA) mechanism for alignment during training. The model supports variable-rate coding across 64 quality levels, achieves a 21.65% BD-Rate saving over VTM-22.0-Intra, uses only 29.9M parameters, and offers fast encoding (61 ms) and decoding (52 ms) speeds.
arXiv:2608.25302v2 Announce Type: replace Abstract: Guided depth super-resolution (GDSR) typically extracts RGB guidance features through convolutional hierarchies, inheriting their fine-to-coarse bi...
WAVE introduces a multi-level discrete wavelet transform (ML‑DWT) to reverse the typical fine‑to‑coarse bias in guided depth super‑resolution. By consuming wavelet sub‑bands and semantic tokens in reverse order, it separates structure and detail reconstruction, applies semantic gating to high‑frequency bands, and fuses modalities via an invertible coupling mechanism. Experiments on multiple benchmarks show that WAVE matches or outperforms existing methods, especially at high upsampling factors where low‑resolution depth has minimal structure.
MambaMPD is a new segmentation framework that leverages Vision Mamba models for marine pollution detection in remote‑sensing imagery. It introduces two structural priors—Frequency‑Aware Augmentation (FAA) and multi‑scale Edge‑Guided Attention (EGA)—to better capture low‑contrast, fragmented pollution patterns and sharpen boundaries. Experiments on the MADOS and M4D datasets show that MambaMPD outperforms existing methods in mIoU while using far less computation than foundation‑model approaches.
arXiv:2509. 05441v4 Announce Type: replace-cross Abstract: Latent generative models compress images into learned embeddings prior to synthesis, and the generation quality critically depends on how faithfully these embeddings preserve visual detail.
FreeTransformSR is a lightweight image super‑resolution network that uses a channel‑wise free low‑rank learnable transform to adaptively modulate features with minimal parameters. It adds a local feature modulation branch with depthwise convolution and a soft complexity adaptive module that fuses local convolution and window self‑attention based on texture characteristics. The model also employs an adaptive intensity modulation strategy and achieves competitive PSNR/SSIM on five benchmark datasets while using only 595K parameters and running faster than competing methods.
Existing object detection methods predominantly utilize sRGB inputs, which are compressed from RAW sensor data using Image Signal Processors (ISP) originally designed for visualization purposes. Compared to RGB images, RAW images possess favorable noise characteristics and richer information representation, which are crucial for object detection, particularly under challenging conditions such as adverse weather or low-light environments.
arXiv:2609.18069v1 Announce Type: new Abstract: Underwater semantic segmentation is essential for marine ecosystem monitoring, yet remains challenging due to severe visual degradation. Light absorpti...
arXiv:2609.39222v1 Announce Type: new Abstract: High-compression tokenizers are essential for scaling latent image generative models. However, aggressive compression creates a fundamental tradeoff be...
The paper proposes a two-stage training strategy for Implicit Neural Representations (INRs) that addresses spectral bias by using a neighbor-aware soft mask to emphasize high-frequency details early in training. In the first stage, the mask assigns higher weights to pixels with strong local variations, encouraging the network to focus on fine edges and textures. The second stage transitions to full-image training, and experiments show consistent improvements in reconstruction quality across existing INR methods.
HyVIC is a configurable variational autoencoder designed for hyperspectral image compression that explicitly separates spatial and spectral feature learning. By allowing independent control of these two aspects, the architecture improves reconstruction fidelity across a wide range of compression ratios, achieving up to 4.66 dB better BD‑PSNR than previous methods. The authors also introduce a metric‑driven strategy for hyperparameter selection and provide code and pretrained models publicly.