Marco Pedersoli

Learn More
Face detection is a mature problem in computer vision. While diverse high performing face detectors have been proposed in the past, we present two surprising new top performance results. First, we show that a properly trained vanilla DPM reaches top performance, improving over commercial and research systems. Second, we show that a detector based on rigid(More)
Weakly supervised object detection, is a challenging task, where the training procedure involves learning at the same time both, the model appearance and the object location in each image. The classical approach to solve this problem is to consider the location of the object of interest in each image as a latent variable and minimize the loss generated by(More)
Motivation: In weakly supervised object detection where only the presence or absence of an object category as a binary label is available for training, the common practice is to model the object location with latent variables and jointly learn them with the object appearance model [1, 5]. An ideal weakly supervised learning method for object detection is(More)
We present a method that can dramatically accelerate object detection with part based models. The method is based on the observation that the cost of detection is likely to be dominated by the cost of matching each part to the image, and not by the cost of computing the optimal configuration of the parts as commonly assumed. Therefore accelerating detection(More)
In this paper we propose a method that aims at automatically editing an image by altering its attributes. More specifically, given an image of a certain class (e.g. a human face), the method should generate a new image as similar as possible to the given one, but with an altered visual attribute (e.g. the same face with a new pose or a different(More)
RCFL versus Casade on VOC2007: plane bike bird boat bottle bus car cat chair cow table dog horse mbike person plant sheep sofa train tv mean speed Exact 24.1 41.3 11.3 3.9 20.8 36.8 35.4 25.5 16.0 19.4 21.2 23.0 42.9 39.8 24.9 14.6 14.3 33.0 22.8 37.4 25.4 1.0 Cascade 24.1 38.7 12.9 3.9 19.9 37.3 35.7 25.9 16.0 19.3 21.2 23.0 40.2 41.5 24.9 14.6 15.1 33.2(More)
In this paper we evaluate the quality of the activation layers of a convolutional neural network (CNN) for the generation of object proposals. We generate hypotheses in a sliding-window fashion over different activation layers and show that the final convolutional layers can find the object of interest with high recall but poor localization due to the(More)
What does this paper demonstrate. We show that a very simple 2D architecture (in the sense that it does not make any assumption or reasoning about the 3D information of the object) generally used for object classification, if properly adapted to the specific task, can provide top performance also for pose estimation. More specifically, we demonstrate how a(More)
In this paper we propose a human detection framework based on an enhanced version of Histogram of Oriented Gradients (HOG) features. These feature descriptors are computed with the help of a precalculated histogram of square-blocks. This novel method outperforms the integral of oriented histograms allowing the calculation of a single feature four times(More)
This paper is focused on the automatic recognition of human events in static images. Popular techniques use knowledge of the human pose for inferring the action, and the most recent approaches tend to combine pose information with either knowledge of the scene or of the objects with which the human interacts. Our approach makes a step forward in this(More)