Detection, segmentation and low-level vision

OD3 — OD3: Optimization-free Dataset Distillation for Object Detection

Salwa K. Al Khatib, Ahmed ElHagry, Shitong Shao et al. · ICLR 2026notableDense predictionpaper ↗code ↗

OD3 distills an object-detection dataset without any bi-level optimization, in two stages it iteratively places real object instances at suitable locations in synthesized images (candidate selection), then screens out low-confidence placements with a pretrained observer model (candidate screening). On MS COCO and PASCAL VOC at compression ratios from 0.25% to 5%, the paper reports new state-of-the-art results, surpassing the prior detection-distillation method by more than 14 points of mAP50 at a 1.0% compression ratio.

Low-Level Dataset Distillation for Medical Image Enhancement

Fengzhi Xu, Ziyuan Yang, Mengyu Sun et al. · arXiv 2025Dense predictionpaper ↗

Argues that low-level tasks (denoising, enhancement) make dataset distillation underdetermined because the many-to-many pixel-level mapping cannot be constrained by a small distilled set the way many-to-one classification labels can; proposes the first low-level DD method, which builds a shared anatomical prior from one representative patient, personalizes it per patient with a Structure-Preserving Personalized Generation module, and injects patient-specific knowledge by matching gradients between networks trained on distilled pairs and on the patient's raw data. Downstream users only receive the abstracted distilled pairs, not raw patient data.

DCOD — Fetch and Forge: Efficient Dataset Condensation for Object Detection

Ding Qi, Jian Li, Jinlong Peng et al. · NeurIPS 2024Dense predictionpaper ↗

Proposes DCOD, the first dataset-condensation framework for object detection, in a two-stage Fetch-and-Forge pipeline: localization and classification information from the real dataset is first embedded into model parameters (Fetch), then synthetic images are reconstructed by inverting that model (Forge); a Foreground-Background Decoupling strategy and an Incremental PatchExpand step increase multi-instance diversity within a fixed storage budget. At a 1% compression rate it reports 46.4% and 24.7% AP50 on VOC and COCO respectively.

OD3 — OD3: Optimization-free Dataset Distillation for Object Detection

Salwa K. Al Khatib, Ahmed ElHagry, Shitong Shao et al. · ICLR 2026notableDense predictionpaper ↗code ↗

OD3 distills an object-detection dataset without any bi-level optimization, in two stages it iteratively places real object instances at suitable locations in synthesized images (candidate selection), then screens out low-confidence placements with a pretrained observer model (candidate screening). On MS COCO and PASCAL VOC at compression ratios from 0.25% to 5%, the paper reports new state-of-the-art results, surpassing the prior detection-distillation method by more than 14 points of mAP50 at a 1.0% compression ratio.

GSDD — GSDD: Generative Space Dataset Distillation for Image Super-resolution

Haiyu Zhang, Shaolin Su, Yu Zhu et al. · AAAI 2024notableDense predictionpaper ↗

GSDD distills a low-resolution/high-resolution image-pair dataset for super-resolution by optimizing codes in the latent space of a pretrained GAN via GAN-inversion, rather than selecting or synthesizing pixels directly, so the stored artifact is a small set of latent codes plus the frozen generator. The paper reports super-resolution performance comparable to prior state-of-the-art distillation methods with about an 8x increase in training efficiency and roughly 93.2% less storage, and shows generalization to real-world degraded images.

UniDD — Towards Universal Dataset Distillation via Task-Driven Diffusion

Ding Qi, Jian Li, Junyao Gao et al. · CVPR 2025notableDense predictionpaper ↗

UniDD extends dataset distillation beyond classification to detection and segmentation with a task-driven diffusion model, first mining task-relevant information by training task-specific proxy models (Universal Task Knowledge Mining), then using those proxies to guide a diffusion process that generates task-specific synthetic images with localized, region-level control rather than the global, single-class updates of prior generation methods (Universal Task-Driven Diffusion). The paper reports consistently outperforming state-of-the-art methods on ImageNet-1K, Pascal VOC and MS COCO across classification, detection and segmentation.

D3S2 — D3S2: Diffusion-Guided Dataset Distillation for Semantic Segmentation

Wenjie Zheng, Haoji Hu, Jiali Lu et al. · arXiv 2026Dense predictionpaper ↗

Proposes D3S2 for dense-prediction dataset distillation in semantic segmentation: a class-balanced set of masks is selected by a greedy strategy that prioritizes underrepresented classes, then a pretrained layout-to-image diffusion model generates images conditioned on those masks, guided at sampling time by a segmentation-consistency loss and a class-wise feature-matching loss for pixel-level alignment; at a 1% compression rate with Mask2Former (Swin-S) it reports 24.99% mIoU on ADE20K and 35.49% on COCO-Stuff, improving over random selection by 9.34 and 5.70 points respectively.