Lihi Zelnik‐Manor
YOU?
Author Swipe
View article: Sharp-It: A Multi-view to Multi-view Diffusion Model for 3D Synthesis and Manipulation
Sharp-It: A Multi-view to Multi-view Diffusion Model for 3D Synthesis and Manipulation Open
Advancements in text-to-image diffusion models have led to significant progress in fast 3D content creation. One common approach is to generate a set of multi-view images of an object, and then reconstruct it into a 3D model. However, this…
View article: FreeAugment: Data Augmentation Search Across All Degrees of Freedom
FreeAugment: Data Augmentation Search Across All Degrees of Freedom Open
Data augmentation has become an integral part of deep learning, as it is known to improve the generalization capabilities of neural networks. Since the most effective set of image transformations differs between tasks and domains, automati…
View article: STMPL: Human Soft-Tissue Simulation
STMPL: Human Soft-Tissue Simulation Open
In various applications, such as virtual reality and gaming, simulating the deformation of soft tissues in the human body during interactions with external objects is essential. Traditionally, Finite Element Methods (FEM) have been employe…
View article: HUGO, a High-Resolution Tactile Emulator for Complex Surfaces
HUGO, a High-Resolution Tactile Emulator for Complex Surfaces Open
Many of our activities rely on tactile feedback perceived through mechanoreceptors in our skin. While visual and auditory devices provide immersive experiences, cutaneous feedback devices are typically limited in the range of sensations th…
View article: Diverse Imagenet Models Transfer Better
Diverse Imagenet Models Transfer Better Open
A commonly accepted hypothesis is that models with higher accuracy on Imagenet perform better on other downstream tasks, leading to much research dedicated to optimizing Imagenet accuracy. Recently this hypothesis has been challenged by ev…
View article: BINAS: Bilinear Interpretable Neural Architecture Search
BINAS: Bilinear Interpretable Neural Architecture Search Open
Practical use of neural networks often involves requirements on latency, energy and memory among others. A popular approach to find networks under such requirements is through constrained Neural Architecture Search (NAS). However, previous…
View article: IQNAS: Interpretable Integer Quadratic Programming Neural Architecture Search
IQNAS: Interpretable Integer Quadratic Programming Neural Architecture Search Open
Realistic use of neural networks often requires adhering to multiple constraints on latency, energy and memory among others. A popular approach to find fitting networks is through constrained Neural Architecture Search (NAS). However, prev…
View article: Multi-label Classification with Partial Annotations using Class-aware Selective Loss
Multi-label Classification with Partial Annotations using Class-aware Selective Loss Open
Large-scale multi-label classification datasets are commonly, and perhaps inevitably, partially annotated. That is, only a small subset of labels are annotated per sample. Different methods for handling the missing labels induce different …
View article: PETA: Photo Albums Event Recognition using Transformers Attention
PETA: Photo Albums Event Recognition using Transformers Attention Open
In recent years the amounts of personal photos captured increased significantly, giving rise to new challenges in multi-image understanding and high-level image understanding. Event recognition in personal photo albums presents one challen…
View article: Semantic Diversity Learning for Zero-Shot Multi-label Classification
Semantic Diversity Learning for Zero-Shot Multi-label Classification Open
Training a neural network model for recognizing multiple labels associated with an image, including identifying unseen labels, is challenging, especially for images that portray numerous semantically diverse labels. As challenging as this …
View article: ImageNet-21K Pretraining for the Masses
ImageNet-21K Pretraining for the Masses Open
ImageNet-1K serves as the primary dataset for pretraining deep learning models for computer vision tasks. ImageNet-21K dataset, which is bigger and more diverse, is used less frequently for pretraining, mainly due to its complexity, low ac…
View article: An Image is Worth 16x16 Words, What is a Video Worth?
An Image is Worth 16x16 Words, What is a Video Worth? Open
Leading methods in the domain of action recognition try to distill information from both the spatial and temporal dimensions of an input video. Methods that reach State of the Art (SotA) accuracy, usually make use of 3D convolution layers …
View article: HardCoRe-NAS: Hard Constrained diffeRentiable Neural Architecture Search
HardCoRe-NAS: Hard Constrained diffeRentiable Neural Architecture Search Open
Realistic use of neural networks often requires adhering to multiple constraints on latency, energy and memory among others. A popular approach to find fitting networks is through constrained Neural Architecture Search (NAS), however, prev…
View article: A Convergence Theory Towards Practical Over-parameterized Deep Neural Networks
A Convergence Theory Towards Practical Over-parameterized Deep Neural Networks Open
Deep neural networks' remarkable ability to correctly fit training data when optimized by gradient-based algorithms is yet to be fully understood. Recent theoretical results explain the convergence for ReLU networks that are wider than tho…
View article: Asymmetric Loss For Multi-Label Classification
Asymmetric Loss For Multi-Label Classification Open
In a typical multi-label setting, a picture contains on average few positive labels, and many negative ones. This positive-negative imbalance dominates the optimization process, and can lead to under-emphasizing gradients from positive lab…
View article: Analysis of subject specific grasping patterns
Analysis of subject specific grasping patterns Open
Existing haptic feedback devices are limited in their capabilities and are often cumbersome and heavy. In addition, these devices are generic and do not adapt to the users' grasping behavior. Potentially, a human-oriented design process co…
View article: Graph Embedded Pose Clustering for Anomaly Detection
Graph Embedded Pose Clustering for Anomaly Detection Open
We propose a new method for anomaly detection of human actions. Our method works directly on human pose graphs that can be computed from an input video sequence. This makes the analysis independent of nuisance parameters such as viewpoint …
View article: Knapsack Pruning with Inner Distillation
Knapsack Pruning with Inner Distillation Open
Neural network pruning reduces the computational cost of an over-parameterized network to improve its efficiency. Popular methods vary from $\ell_1$-norm sparsification to Neural Architecture Search (NAS). In this work, we propose a novel …
View article: BRML Grasp Dataset
BRML Grasp Dataset Open
The dataset contains kinematic and kinetic measurements of 31 subjects grasping five different objects multiple times (1083 grasp instances).
View article: Dynamic-Net: Tuning the Objective Without Re-Training for Synthesis Tasks
Dynamic-Net: Tuning the Objective Without Re-Training for Synthesis Tasks Open
One of the key ingredients for successful optimization of modern CNNs is identifying a suitable objective. To date, the objective is fixed a-priori at training time, and any variation to it requires re-training a new network. In this paper…
View article: Adversarial Feedback Loop
Adversarial Feedback Loop Open
Thanks to their remarkable generative capabilities, GANs have gained great popularity, and are used abundantly in state-of-the-art methods and applications. In a GAN based model, a discriminator is trained to learn the real data distributi…
View article: XNAS: Neural Architecture Search with Expert Advice
XNAS: Neural Architecture Search with Expert Advice Open
This paper introduces a novel optimization method for differential neural architecture search, based on the theory of prediction with expert advice. Its optimization criterion is well fitted for an architecture-selection, i.e., it minimize…
View article: Is Image Memorability Prediction Solved?
Is Image Memorability Prediction Solved? Open
This paper deals with the prediction of the memorability of a given image. We start by proposing an algorithm that reaches human-level performance on the LaMem dataset - the only large scale benchmark for memorability prediction. The sugge…
View article: ASAP: Architecture Search, Anneal and Prune
ASAP: Architecture Search, Anneal and Prune Open
Automatic methods for Neural Architecture Search (NAS) have been shown to produce state-of-the-art network models. Yet, their main drawback is the computational complexity of the search process. As some primal methods optimized over a disc…
View article: Dynamic-Net: Tuning the Objective Without Re-training for Synthesis\n Tasks
Dynamic-Net: Tuning the Objective Without Re-training for Synthesis\n Tasks Open
One of the key ingredients for successful optimization of modern CNNs is\nidentifying a suitable objective. To date, the objective is fixed a-priori at\ntraining time, and any variation to it requires re-training a new network. In\nthis pa…
View article: Dynamic-Net: Tuning the Objective Without Re-training.
Dynamic-Net: Tuning the Objective Without Re-training. Open
One of the key ingredients for successful optimization of modern CNNs is identifying a suitable objective. To date, the objective is fixed a-priori at training time, and any variation to it requires re-training a new network. In this paper…
View article: Learning to Maintain Natural Image Statistics
Learning to Maintain Natural Image Statistics Open
Maintaining natural image statistics is a crucial factor in restoration and generation of realistic looking images. When training CNNs, photorealism is usually attempted by adversarial training (GAN), that pushes the output images to lie o…
View article: Saliency Driven Image Manipulation
Saliency Driven Image Manipulation Open
Have you ever taken a picture only to find out that an unimportant background object ended up being overly salient? Or one of those team sports photos where your favorite player blends with the rest? Wouldn't it be nice if you could tweak …