Every image is encoded using an ImageNet pre-trained ResNet18 to a latent feature map z R4 4 256. A flattened version of z is processed with one linear layer with output channels equal to N 3to get the predictions for points u and visibility v. We apply the sigmoid function to the visibility predictions v to enforce a numerical range [0,1]. Our models are trained using Adam optimizer with learning rate equal to 1e-4. In detail, scale is sampled from the range [0.7, 1.2], vertical translation is up to 38 pixels and we also apply 2D rotation up to 40 degrees. For camera equivariance the image is simply flipped horizontally and given as input to the network to estimate the pose.

A.1 Network Architecture Here, we describe the architecture of the eVAE presented in Figure 1 of the main paper, in more detail. Event Context Network: We adapt the architecture proposed in [21] for the event context network, but without the feature transformation preprocessing steps. In our implementation, we use three Conv1d layers of 64, 128 and 1024 channels each followed by BatchNorm and a ReLU activation. At the end of the ECN, we add the temporal features (see Appendix A.2) to the N 1024 feature tensor, and execute the max operation to result in a context vector. The sizes of the intermediate features and the context feature are hyperparameters that can be varied based on the application, data complexity etc. Encoder: The encoder for the VAE is composed of two layers, of sizes 1024 and 256 respectively, resulting in two output vectors of 1 8 each, corresponding to the mean and standard deviation for the latent space vector.

One of the dominant paradigms in self-supervised learning (SSL), illustrated by MoCo or DINO, aims to produce robust representations by capturing features that are insensitive to certain image transformations such as illumination, or geometric changes. This strategy is appropriate when the objective is to recognize objects independently of their appearance. However, it becomes counterproductive as soon as appearance itself constitutes the discriminative signal. In weather analysis, for example, rain streaks, snow granularity, atmospheric scattering, as well as reflections and halos, are not noise: they carry the essential information. In critical applications such as autonomous driving, ignoring these cues is risky, since grip and visibility depend directly on ground conditions and atmospheric conditions. We introduce ST-STORM, a hybrid SSL framework that treats appearance (style) as a semantic modality to be disentangled from content. Our architecture explicitly separates two latent streams, regulated by gating mechanisms. The Content branch aims at a stable semantic representation through a JEPA scheme coupled with a contrastive objective, promoting invariance to appearance variations. In parallel, the Style branch is constrained to capture appearance signatures (textures, contrasts, scattering) through feature prediction and reconstruction under an adversarial constraint. We evaluate ST-STORM on several tasks, including object classification (ImageNet-1K), fine-grained weather characterization, and melanoma detection (ISIC 2024 Challenge). The results show that the Style branch effectively isolates complex appearance phenomena (F1=97% on Multi-Weather and F1=94% on ISIC 2024 with 10% labeled data), without degrading the semantic performance (F1=80% on ImageNet-1K) of the Content branch, and improves the preservation of critical appearance

Fat is key for fudgy brownies. More information Adding us as a Preferred Source in Google by using this link indicates that you would like to see more of our content in Google News results. Breakthroughs, discoveries, and DIY tips sent six days a week. Astronauts aboard the International Space Station have brownies on their menu too . But what makes a perfect brownie?