The paper proposes a method for adapting pretrained latent spaces without retraining the encoder or decoder by inserting an invertible bi‑Lipschitz adapter that aligns the latent coordinates to a semantic teacher. This offline alignment preserves the compression and reconstruction properties of the original autoencoder while improving semantic organization. Experiments on two medical imaging datasets show that constrained adaptation enhances feature metrics, seed stability, and counterfactual edit strength, and remains competitive on downstream classification tasks.
arXiv:2512. 09185v4 Announce Type: replace-cross Abstract: Understanding disease progression is a central clinical challenge with direct implications for early diagnosis and personalized treatment.
arXiv:2608. 08135v1 Announce Type: cross Abstract: Cross-modality medical image translation can reduce the burden of multi-modal acquisitions, yet the field remains constrained by two coupled limitations: methods operate on 2D slices or 3D patches rather than whole volumes, and train a separate model for each translation task.
arXiv:2608. 05683v1 Announce Type: cross Abstract: Cross-modal alignment of visual and textual representations is fundamental to multimodal medical image understanding, yet remains hindered by uncertainty in both modalities under real-world clinical conditions.
arXiv:2609.36348v1 Announce Type: cross Abstract: Generative and representation learning remain asymmetrically connected: semantic representations are used to improve diffusion generation, whereas th...
arXiv:2505. 04486v4 Announce Type: replace-cross Abstract: Flow matching models have shown great potential in image generation tasks among probabilistic generative models.
arXiv:2608.22532v1 Announce Type: new Abstract: Domain shift across imaging modalities and acquisition sites remains a significant barrier to the clinical deployment of segmentation models. Source-fr...
CoVA introduces a lightweight framework that composes VAE features online to provide stage‑wise guidance for flow‑matching training, improving semantic discrimination and spatial structure without extra inference overhead. By jointly training a small latent composer with class supervision, masked reconstruction, and flip consistency, CoVA aligns intermediate SiT features early and later uses detached composer posteriors for sample‑dependent soft‑label conditioning. On ImageNet 256×256, CoVA cuts SiT‑XL/2 FID from 17.2 to 7.4 in 400K steps and achieves an FID of 2.03 in only 400 epochs with classifier‑free guidance, adding just 8.5 M training‑only parameters.
arXiv:2511.05782v3 Announce Type: replace Abstract: Unsupervised domain adaptation (UDA) for medical image segmentation remains challenging due to substantial domain shifts across imaging modalities,...
The paper introduces SCION, a single-model framework that jointly learns generative and self‑supervised representation tasks without relying on labels or pretrained encoders. SCION uses a pixel‑space encoder conditioned on a flow timestep and a shared embedding, training contrastive and flow‑matching objectives simultaneously with gradient‑norm balancing and stop‑gradient techniques. On ImageNet 256×256, SCION achieves state‑of‑the‑art FID scores (8.92 with JiT‑B and 5.89/3.47 with JiT‑L) surpassing prior class‑unconditional and pretrained‑conditioned methods.
arXiv:2605. 08398v2 Announce Type: replace Abstract: In this work, we show that Latent Flow-Matching (LFM) models are robust to different types of perturbations, including data reduction and model capacity shrinkage.
The paper introduces a 3D-CLIP encoder trained with structured hard negatives to improve vision‑language alignment for text‑to‑CT generation. This encoder drives a latent diffusion model that operates directly in 3D latent space, eliminating spatial artifacts from super‑resolution pipelines. Experiments on the CT‑RATE dataset show state‑of‑the‑art image fidelity and factual correctness across 18 pathological conditions, with lower inference time and GPU memory usage than competing methods.
FlowMoDL is an unrolled neural network designed for highly accelerated 4D flow MRI reconstruction, optimizing both anatomical magnitude and phase-derived velocity accuracy. It alternates a learned (3+1)D spatiotemporal denoiser with conjugate‑gradient data‑consistency updates, using a dual‑pathway conditioning scheme to handle acceleration factors from 10× to 50×. Trained with a deep‑supervision composite loss that penalizes velocity magnitude and angular errors, FlowMoDL outperforms classical and deep‑learning baselines on the multi‑center CMRx4DFlow dataset, achieving superior gradient‑step efficiency and robust convergence across all acceleration factors.