Organizations: School of Media Engineering, Communication University of Zhejiang, Hangzhou 310018, China · Samsung Research America, Irvine, CA 92612, USA · School of Computer Science and Engineering, The University of New South Wales, Sydney, NSW 2033, Australia
Image retargeting aims to resize images while preserving salient content free from distortion and stretching. Recent learning-based solutions reduce reliance on traditional hand-crafted features and optimization procedures, leading to more efficient and generalizable performance. Nevertheless, existing approaches still struggle to preserve global structure and content fidelity. To address these limitations, we propose AttenIR, an unsupervised image retargeting framework requiring no manual labels, based on content-aware mesh warping, which ef- fectively couples dense visual attention with sparse mesh constraints. We for- mulate the mesh deformation as a vertex displacement regression task. Guided by saliency priors, our method jointly optimizes scale, structure, and geometry preservation losses. By combining dense visual attention with semantic features, it achieves robust, unsupervised retargeting without manual annotations. Our primary contributions include (1) a lightweight unsupervised method for image retargeting that ensures content fidelity and preserves global structure, and (2) a coarse-to-fine projection strategy to align dense visual attention priors with sparse mesh constraints, enabling content-aware deformation at the pixel level. Experiments demonstrate that our method achieves competitive and superior performance compared with both traditional and state-of-the-art learning-based methods.
Figures & tables
Figure 1: Overall network architecture of the proposed method. The network takes an image and a target aspect ratio as input and outputs the vertex displacement of the original rigid mesh, which is then used to warp the input image and produce the retargeting result.
Figure 2: Comparison between results generated by cell-level coarse constraints and pixel-level fine-grained constraints. The zoom-in views indicate that methods employing cell-level constraints, obtained by averaging dense saliency values over each mesh cell, cannot adequately preserve appearance and structural features. In contrast, our pixel-level constraints yield more favorable results.
Figure 3: Flowchart of building the pixel-wise constraints. The cell-level mesh loss is first transformed into vertex-level values by averaging the losses over adjacent cells. These vertex values are subsequently interpolated within each deformed cell to derive candidate pixel-level constraints, and a soft mask is employed to select valid cell contributions.
Figure 4: Illustration of the soft mask Aij(p) . For a grid cell Cij , the normalized minimum signed distance is defined as dˉij(p)=mint∈{0,1,2,3}∥et∥2et×ep , where et denotes an oriented cell edge, and ep denotes the vector connecting the starting vertex of et to pixel p . Pixels inside and outside the cell are assigned positive and negative signed distances, respectively. The Sigmoid function maps dˉij(p) to the corresponding soft mask weight Aij(p) .
Figure 5: Mesh geometry preservation constraint, which consists of intra-cell and inter-cell terms.
Methods
0.5 ×
0.75 ×
1.25 ×
1.5 ×
1.75 ×
CR
0.4652/0.7761
0.3251/0.5347
–/–
–/–
–/–
SCL
0.6357/0.7721
0.3074/0.3507
0.2873/0.2531
0.5323/0.5392
0.8096/0.7096
SC [ 1 ]
0.4054/0.5615
0.2984/ 0.2944
0.2597/ 0.2182
0.3352 /0.4945
0.6042/0.6991
SNS [ 9 ]
0.3280 /0.6689
0.2941/0.4677
0.2550/0.3993
0.4901/0.4971
0.4824/ 0.4853
WSSDCNN [ 6 ]
0.7934/0.7304
0.5268/0.4249
–/–
–/–
–/–
GDPM [ 11 ]
0.8658/0.8158
0.6828/0.7113
0.4878/0.4393
0.4974/0.6837
0.4572 /0.6243
Table 1: Quantitative comparisons with representative retargeting approaches on two datasets using the object-based distortion error metric from Object-IR [ 22 ] , including CR (Cropping), SCL (Simple Scaling), SC [ 1 ] , SNS [ 9 ] , WSSDCNN [ 6 ] , GDPM [ 11 ] , Object-IR [ 22 ] . Each cell is formatted as “value on ObjectIR testing dataset [ 22 ] / value on RetargetMe dataset [ 31 ] ”. Red and Cyan colorboxes indicate the best performance on the ObjectIR [ 22 ] and RetargetMe [ 31 ] datasets, respectively. Underlined values denote the second-best results, and lower values indicate superior retargeting quality.
Methods
DISTS ↓
MUSIQ ↑
Prune [ 34 ]
0.2016
65.56
Ours
0.1429
70.79
Table 2: Additional perceptual quality comparison with PruneRepaint [ 34 ] on Retargetme [ 31 ] dataset. DISTS and MUSIQ are used to measure perceptual distortion and no-reference image quality, respectively.
Figure 6: Visual comparison with PruneRepaint [ 34 ] . DISTS and MUSIQ are reported below each retargeted result. Lower DISTS and higher MUSIQ values indicate better perceptual consistency and overall image quality, respectively.
Figure 7: Visual comparison with the SOTA retargeting method [ 22 ] across different scales , where colored boxes mark the distorted artifacts of [ 22 ] .
Figure 8: More visual comparisons on the RetargetMe dataset [ 31 ] . All results are generated scaling the width to 0.5 × its original value. From left to right are the original input, CR (manual cropping), SCL (simple scaling), SNS [ 38 ] , WSSCNN [ 6 ] , CycleIR [ 37 ] , GPDM [ 11 ] , ObjectIR [ 22 ] and our method.
Figure 9: Comparison on challenging examples with complex backgrounds (1–3 lines) and dispersed foreground regions (the last line) .
Figure 10: Visual comparisons of ablation analysis on loss functions, mesh resolution, as well as pixel-level and cell-level constraints.
ID
Loss function
Projection
Mesh resolution
Dataset
Lc
Lt
Lg
Cell
Pixel
5×5
11×11
22×22
ObjectIR [ 22 ]
RetargetMe [ 31 ]
1
✓
✓
✓
✓
0.6391
0.7487
2
✓
✓
✓
✓
0.4032
0.5493
3
✓
✓
✓
✓
0.9876
1.0000
4
✓
✓
✓
✓
✓
0.3096
0.4568
5
✓
✓
✓
✓
✓
0.2917
0.4420
Table 3: Ablation studies on the ObjectIR [ 22 ] and RetargetMe [ 31 ] datasets. The gray row corresponds to the results of our model, while the boldfaced and underlined entries indicate the best and second-best results, respectively.
Metric
Ours
Prune [ 34 ]
SC [ 1 ]
ObjectIR [ 22 ]
Avg.
3.1040
2.7560
2.0240
2.3760
Var.
1.0811
1.6383
1.2617
1.0998
Table 4: User Study on the RetargetMe Dataset [ 31 ] . 25 participants were invited to score 30 image groups randomly sampled from the RetargetMe dataset and generated by representative retargeting algorithms [ 34 , 1 , 22 ] . The mean score and variance for each method are reported in the table.
Transformer-based models have made remarkable progress in image restoration (IR) tasks. However, the quadratic complexity of self-attention in Transformer hinders its applicability to high-resolution images. Existing methods mitigate this issue with sparse or window-based attention, yet inherently limit global context modeling. Linear attention, a variant of softmax attention, demonstrates promise in global context modeling while maintaining linear complexity, offering a potential solution to the above challenge. Despite its efficiency benefits, vanilla linear attention suffers from a significant performance drop in IR, largely due to the low-rank nature of its attention map. To counter this, we propose Rank Enhanced Linear Attention (RELA), a simple yet effective method that enriches feature representations by integrating a lightweight depthwise convolution. Building upon RELA, we propose an efficient and effective Vision Transformer, named LAformer. LAformer eliminates hardware-inefficient operations such as softmax and window shifting, enabling efficient processing of high-resolution images. Extensive experiments across 7 IR tasks and 21 benchmarks demonstrate that LAformer outperforms SOTA methods and offers significant computational advantages. Furthermore, we extend LAformer to diffusion-based and flow-based visual generation, showcasing its strong potential as a competitive alternative to DiT and SiT. Code and models are available at https://github.com/shallowdream204/LAformer.
Yuang Ai
University of Chinese Academy of Sciences, Beijing, China
In this work, we propose Image-to-Image Rectified Flow Reformulation (I2I-RFR), a practical plug-in reformulation that recasts standard I2I regression networks as continuous-time transport models. While pixel-wise I2I regression is simple, stable, and easy to adapt across tasks, it often over-smooths ill-posed and multimodal targets, whereas generative alternatives often require additional components, task-specific tuning, and more complex training and inference pipelines. Our method augments the backbone input by channel-wise concatenation with a noise-corrupted version of the ground-truth target and optimizes a simple t-reweighted pixel loss. This objective admits a rectified-flow interpretation via an induced velocity field, enabling ODE-based progressive refinement at inference time while largely preserving the standard supervised training pipeline. In most cases, adopting I2I-RFR requires only expanding the input channels, and inference can be performed with a few explicit solver steps (e.g., 3 steps) without distillation. Extensive experiments across multiple image-to-image translation and video restoration tasks show that I2I-RFR generally improves performance across a wide range of tasks and backbones, with particularly clear gains in perceptual quality and detail preservation. Overall, I2I-RFR provides a lightweight way to incorporate continuous-time refinement into conventional I2I models without requiring a heavy generative pipeline.
Motion retargeting transfers motion across characters with different skeletal structures while preserving semantic intent and physical plausibility. Despite recent progress, two fundamental questions remain: (i) how can reliable source-motion semantics be learned without high-quality paired retargeting data, and (ii) how should retargeting be formulated when no reliable paired motion can serve as a definitive regression objective? Existing methods commonly preserve semantics by constraining predictions toward copied motions. However, such initializations entangle useful articulation cues with artifacts caused by mismatched skeletal proportions and body geometry. Moreover, directly regressing a final motion in one forward pass is restrictive because retargeting is inherently underdetermined, and the desired solution must balance semantic fidelity with target-specific physical and temporal constraints rather than match a unique paired target. Motivated by these limitations, we propose ReFM, a source-mesh-agnostic, energy-guided model that reformulates motion retargeting as progressive refinement. First, an SO(3) canonicalizer removes redundant global-orientation variations. Second, a cross-character semantic encoder, pretrained through contrastive learning, provides a character-invariant representation for both optimization guidance and semantic evaluation. ReFM then progressively refines an initialized target motion through a learned flow guided by semantic consistency, physical plausibility, temporal coherence, and minimal motion modification. The framework is compatible with different initialization strategies, including both direct motion copying and Autodesk HumanIK, an industry-standard full-body inverse-kinematics retargeting system.