Search

In this article we present a new appearance-based approach for the classification and the localization of 3-D objects in complex scenes. A main problem for object recognition is that the size and the appearance of the objects in the image vary for 3-D transformations. For this reason, we model the region of the object in the image as well as the object features themselves as functions of these transformations. We integrate the model into a statistical framework, and so we can deal with noise and illumination changes. To handle heterogeneous background and occlusions, we introduce a background model and an assignment function. Thus, the object recognition system becomes robust, and a reliable distinction, which features belong to the object and which to the background, is possible. Experiments on three large data sets that contain rotations orthogonal to the image plane and scaling with together more than 100000 images show that the approach is well suited for this task.

Given an image sequence recorded by a hand-held camera we examine the computation of a light field without any further input data. Using structure-from-motion algorithms and optimization techniques camera motion and a 3-D reconstruction of the scene are established. The light field is completed by computing local depth information for each input image. During experimental evaluation a special focus is set on the effects of falsely estimated intrinsic parameters as well as different depth representations on the quality of the resulting light fields.

Objective knowledge of tissue density distribution in CT/MRI brain datasets can be related to anatomical or neuro-functional regions for assessing pathologic conditions characterised by slight differences. The process of monitoring illness and its treatment could be then improved by a suitable detection of these variations. In this paper, we present an approach for three-dimensional (3D) classification of brain tissue densities based on a hierarchicalartificial neural network(ANN) able to classify the single voxels of the examined datasets. The method developed was tested on case studies selected by an expert neuro-radiologist and consisting of both normal andpathological conditions. The results obtained were submitted for validation to a group of physicians and they judged the system to be really effective in practical applications.

Due to the loss of range information, projections as input data for a 3-D object recognition algorithm are expected to increase the computational complexity. In this work, however, we demonstrate that this deficiency carries potential for complexity reduction of major vision problems. We show that projections provide a reduction of feature dimensions, and lead to structures exhibiting simple combinatorial properties. The theoretical framework is embedded in a probabilistic setting which deals with uncertainties and variations of observed features. In statistics marginal densities and the assumption of independency prove to be the key tools when one encounters projections. The examples discussed in this paper include feature matching, pose estimation as well as classification of 3-D objects. The final experimental evaluation demonstrates the practical importance of the marginalization concept and independency assumptions.

In this paper, we present an integrated approach for recognizing both the word sequence and the syntactic–prosodic structure of a spontaneous utterance. The approach aims at improving the performance of the understanding component of speech understanding systems by exploiting not only acoustic–phonetic and syntactic information, but also prosodic information directly within the speech recognition process. Whereas spoken utterances are typically modelled as unstructured word sequences in the speech recognizer, our approach includes phrase boundary information in the language model and provides HMMs to model the acoustic and prosodic characteristics of phrase boundaries. This methodology has two major advantages compared to purely word-based speech recognizers. First, additional syntactic–prosodic boundaries are determined by the speech recognizer which facilitates parsing and resolve syntactic and semantic ambiguities. Second – after having removed the boundary information from the result of the recognizer – the integrated model yields a 4% relative word error rate (WER) reduction compared to a traditional word recognizer. The boundary classification performance is equal to that of a separate prosodic classifier operating on the word recognizer output, thus making a separate classifier unnecessary for this task and saving the computation time involved. Compared to the baseline word recognizer, the integrated word-and-boundary recognizer does not involve any computational overhead.

In this paper we present a neural network (NN) based system for recognition and pose estimation of 3D objects from a single 2D perspective view. We develop an appearance based neural approach for this task. First the object is represented in a feature vector derived by a principal component network. Then a NN classifier trained with R esilient back prop agation (Rprop) algorithm is applied to identify it. Next pose parameters are obtained by four NN estimators trained on the same feature vector. Performance on recognition and pose estimation for real images under occlusions are shown. Comparative studies with two other approaches are carried out.

Bovine morulae and blastocysts were either produced in vitro through maturation, fertilization and culture of immature oocytes recovered from slaughterhouse-derived ovaries, collected in vivo or obtained after 24 h in vitro culture of in vivo collected embryos. The morulae and blastocysts were classified into four categories of embryo quality and two stages of embryonic development. Embryos were frozen by a controlled freezing method using 10% glycerol as a cryoprotectant. The ability of individual embryos to withstand freeze/thawing was measured immediately before and after cryopreservation by changes in CO2 production from (U-14C)glucose during a 2 h incubation period in a non-invasive closed system immediately before and after cryopreservation. Post-thaw survival was assessed by development in vitro during a 48 h culture period. Survival rates and oxidative metabolism after freeze/thawing were similar in embryos of the two developmental stages. However, after freeze/thawing, the rate of CO2 production of in vitro produced embryos was reduced to one half of their pre-freeze levels and was associated with poor survival rates. In vivo collected embryos had a significantly better tolerance to freezing and higher survival rates. However, when in vivo embryos were exposed to in vitro culture conditions, the rates of CO2 production and survival were significantly reduced. Pre-freeze embryo quality affected post-thaw in vitro development and metabolic activity markedly in embryos produced in vitro or pre-exposed to in vitro culture conditions. While there was no relationship between pre-freeze levels of CO2 production and post-thaw in vitro embryo development, all embryos which developed in vitro after freezing/thawing retained at least 58% of the pre-freeze levels of CO2 production regardless of their origin. Results of the present study indicate that embryos produced in vitro or pre-exposed to in vitro culture conditions are more sensitive to cryo-injury. This sensitivity is affected by embryo quality and is similarly reflected at the biochemical level. Determination of oxidative metabolism offers a feasibility for selection of viable morulae/blastocysts after freezing/thawing.

In this paper we discuss and compare di!erent approaches to appearance-based object recognition and pose estimation. Images are considered as high-dimensional feature vectors which are transformed in various manners: we use di!erent types of non-linear image-to-image transforms composed with linear mappings to reduce the feature dimensions and to beat the curse of dimensionality. The transforms are selected such that special objective functions are optimized and available image data provide some invariance properties. The paper mainly concentrates on the comparison of preprocessing operations combined with di!erent linear projections in the context of appearance-based object recognition. The experimental evaluation provides recognition rates and pose estimation accuracy. ( 1999 Pattern Recognition Society. Published by Elsevier Science Ltd. All rights reserved.

This study examined the effects of adding a macromolecule, polyvinylpyrrolidone (10% PVP) and a sugar (0.3 M trehalose) to vitrification solutions (VS) containing either one (40% ethylene glycol [EG], two (25% EG+25% DMSO) or three (20% EG+20% DMSO+10% 1, 3-butanediol [BD]) permeable cryoprotectants on the survival and hatching of IVP bovine embryos, following vitrification, warming and in-straw cryoprotectant dilution. Grade 1 and 2 compact morulae and blastocysts were selected on Day 7 (Day 0=IVF) of culture in SOFaaBSA and equilibrated for 10 min at room temperature in 10% EG. Following exposure, for up to 1 min at 4 degrees C, to one of the above VS (with or without PVP+trehalose), the embryos were loaded into straws and immersed in liquid nitrogen. Following warming and in-straw cryoprotectant dilution, the embryos were cultured for 48 h to assess hatching. There was no effect of VS on the survival of embryos after 24 h, however fewer compact morulae than blastocysts survived after 24 h (24% vs. 75%; P

The purpose of our experiments was to explore the effects of platelet-derived growth factor (PDGF)-supplementation at the various steps of in vitro production of bovine embryos using protein-free media. Cumulus-oocyte-complexes (COC) were collected by slicing abattoir ovaries and then dividing the COC into 2 morphological categories. After maturation for 24 h in TCM-199 supplemented with hormones and either 20% estrous cow serum (ECS) or 1 mg/ml polyvinyl-alcohol (PVA), oocytes were co-incubated for 19 h with frozen/thawed spermatozoa from bull of proven fertility. The semen was diluted in Fert-Talp supplemented with heparin, hypotaurine and epinephrine and either 6 mg/ml bovine serum albumin (BSA) or 1 mg/ml PVA. Presumptive zygotes were transferred into embryo culture medium containing either 20% ECS or 1 mg/ml PVA for a total of 10 d. The PDGF was added at concentrations of 1, 10 or 100 ng/ml to the maturation medium (Experiment 1), fertilization medium (Experiment 2) or culture medium from Day 1 on (Experiment 3), respectively, or at 1 ng/ml PDGF to both the fertilization and culture medium from Day 3 on (Experiment 4), with each medium supplemented with PVA. Oocytes/embryos incubated in the absence of PDGF in media supplemented with either ECS or PVA served as controls. An average of 20 COC was incubated in 1 droplet under silicone oil, and each experiment contained 4 to 6 replicates. No significant differences were found among the various concentrations of PDGF, nor did PDGF-supplementation during maturation (Experiment 1) or embryo culture on Day 1 (Experiment 3) significantly affect development of oocytes/embryos (34.7 +/- 3.5 to 40.4 +/- 2.5% morulae, 11.9 +/- 2.4 to 18.8 +/- 2.5% blastocysts; and 23.2 +/- 2.3 to 27.5 +/- 3.4% morulae, 11.5 +/- 2.6 to 12.7 +/- 2.3% blastocysts, respectively; x +/- SEM). In the presence of 10 ng/ml PDGF in the fertilization medium development to morulae and blastocysts was similar to that of the ECS-group, and was higher (P0.05) than that of the PVA-control (ECS: 32.1 +/- 4.6 and 13.8 +/- 2.7%; PVA: 17.5 +/- 0.8 and 6.1 +/- 1.3%; PDGF: 30.6 +/- 3.0 and 14.0 +/- 2.2%, respectively). Development to morulae/blastocysts was increased, and was at the same level as in the ECS-group when the fertilization and/or embryo culture medium on Day 3 contained PDGF compared with the PVA-control group (morulae: ECS 25.3 +/- 4.4%, PVA 13.9 +/- 2.2% [P0.05], PDGF 16.7 +/- 3.2 to 19.1 +/- 1.1%; blastocysts: ECS 5.3 +/- 2.1%, PVA 5.0 +/- 1.7%, PDGF 7.1 +/- 1.6 to 9.1 +/- 1.7%, respectively). These results indicate that under our laboratory conditions PDGF can elevate low rates of development and the addition of PDGF to the fertilization medium enhances bovine preimplantation embryonic development. Thus, PDGF can be potentially an important factor in a completely defined medium to substitute the effects of serum.

This study examined the role of protein supplementation at the various steps of the in vitro production of bovine embryos derived from two different morphological categories of COC. The basic medium was TCM 199 and was supplemented with hormones during maturation in vitro and either estrous cow serum (ECS), bovine serum albumin (BSA) at various concentrations or polyvinyl-alcohol (PVA). Fertilization in vitro was carried out using frozen-thawed semen or one bull in Fert-talp containing heparin, hypotaurin and epinephrine and either 6 mg/ml BSA or 1 mg/ml PVA. In vitro culture up to the blastocyst stage was performed in TCM 199 supplemented with either ECS, BSA or PVA. The first experiment investigated the influence of different medium-supplements (ECS, BSA or PVA) on nuclear maturation and revealed no significant differences among treatment groups nor between categories of COC (63.9% to 74.9% and 48.9% to 77.0%, respectively). The time course of in vitro fertilization was elucidated in Experiment 2 in medium supplemented with either protein or PVA during maturation and fertilization. Penetration was not affected (70.9% to 79.3% penetration 12 h after onset of oocyte-sperm-co-incubation), but formation of pronuclei was decreased (P0.05) 12 and 19 h after onset of oocyte-sperm-co-incubation and was retarded in medium supplemented with PVA (12 h: 63.8 vs 21.4 %; 19 h: 57.5 vs 20.8 %, respectively) while cleavage was not affected. In Experiment 3, six treatment groups were formed in which the two different morphological categories of cumulus-oocyte-complexes (COC) were incubated in basic medium supplemented with 1) ECS during maturation and embryo culture and BSA during fertilization; 2) PVA during maturation and embryo culture, fertilization medium with PVA; 3) PVA during maturation and embryo culture, fertilization medium with BSA; 4) BSA (1 mg/ml) during maturation, fertilization and embryo culture; 5) BSA (6 mg/ml) during maturation, fertilization and embryo culture; and 6) BSA (10 mg/ml) during maturation, fertilization and embryo culture. The rates of cleavage and the development to morulae or blastocysts did not differ (P0.05) among treatment groups and between both categories of COC and were showing a high degree of variability (cleavage 54.0% to 65.1% and 41.3% to 55.7%, respectively; morulae 25.3% to 53.0% and 26.0% to 51.2%, respectively; blastocysts 5.4% to 24.7% and 0.6% to 20.3%, respectively). Parthenogenetic activation only rarely occurred in medium containing PVA throughout all steps of in vitro production of bovine embryos (Experiment 4) and led to early cleavage stages (8%), but no development to morula- or blastocyst-stages was observed. It is concluded that 1) formation of pronuclei was retarded in medium lacking protein-supplementation, indicating that BSA is required for regular fertilization in vitro and 2) under our experimental conditions, protein-supplementation is not necessary for maturation and development up to the blastocyst stage in vitro.

Ultrasound-guided follicular aspiration was performed on 29 Holstein-Friesian cows/heifers twice weekly at 3- to 4-d intervals over a period of 2 consecutive estrous cycles (total 42 d). For visualization of the ovaries and guidance of the aspiration needle, a 6.5 MHz fingertip probe on a 62 cm probe carrier was inserted into the vagina. The disposable aspiration needle was connected to a permanent rinse tubing system, thus ensuring minimum death of oocytes in the aspiration processs. After penetration of the vaginal wall, the needle was inserted into a follicle of the rectally fixed ovary. Cumulus oocyte complexes (COC) were aspirated at a pressure of 100 mm Hg. In the first experiment, the effect of an additional gonadotropin treatment 4 d prior to aspiration was investigated in 8 lactating cows. Following FSH-treatment, the number of aspirated follicles was higher (P < 0.05) than in the nontreated animals (10.6 +/- 0.7 vs 8.9 +/- 0.5). The number of recovered COC (7.0 +/- 0.6 vs 5.8 +/- 0.5), the recovery rate (COC per aspirated follicle) (66.6% vs 65.4%), the percentage of viable COC (56.8% vs 52.1%), the cleavage rate upon in vitro maturation and in vitro fertilization (56.7% vs 59.8%) as well as the rate of morula/blastocyst formation (3.8% vs 2.9%) were similar in both groups. In the second experiment, follicles were aspirated in 4 lactating cows, 6 dry cows, 4 pregnant cows (first 35 d of pregnancy), and 4 heifers. The average number of aspirated follicles and recovered COC was higher (P < 0.05) in the first 2 groups (10.6 +/- 0.6 and 9.3 +/- 0.7 follicles; 7.2 +/- 0.5 and 6.9 +/- 0.7 oocytes) than in trie 2 other treatment groups (7.3 +/- 0.5 and 8.1 +/- 0.5 follicles; 5.0 +/- 0.4 and 5.7 +/- 0.5 oocytes). The percentage of viable COC was higher (P < 0.05; 68.3%) in lactating animals than in all the other groups (49.7, 52.5 and 57.4%, respectively). Similarly, upon in vitro fertilization, cleavage rate was higher (P < 0.05; 63.4%) in lactating cows than in the other groups (43.7, 50.5, 55.1%, respectively). A total of 21.5, 22.7, 11.9 and 13.5%, respectively, in the 4 groups of the in vitro fertilized oocytes reached the morula and blastocyst stages. After transfer of a total of 48 embryos 22 pregnancies (45.8%) were established as detected on Day 65. We conclude that 1) repeated aspiration of viable COC at short intervals is possible, 2) additional FSH-treatment does not increase oocyte yields, and 3) viable blastocysts can be produced from cattle at various reproductive phases irrespective of the reproductive phase.

The mammary gland of transgenic livestock has recently been shown to be an appropriate site for the production of recombinant human pharmaceutical proteins in biologically active forms. The purpose of our experiments was to generate transgenic sheep in which the human blood clotting factor VIII gene (hFVIII) is expressed in the ovine mammary gland. A total of 920 zygotes was recovered in 119 surgical collections from superovulated donors with a recovery rate of 83.4% (zygotes per corpora lutea). A total of 471 zygotes were microinjected with either the metallothionein human blood clotting factor VIII cDNA (MT-FVIII) or β-lactoglobulin FVIII cDNA (β-Lac VFIII). One to four microinjected zygotes were transferred to 130 recipients of which 76 (58.5%) were diagnosed as pregnant and delivered a total of 124 lambs. Pregnancy rate was significantly increased when four zygotes had been transferred per recipient (70.8% vs. 47.8% to 58.4%). Three transgenic lambs were identified by southern blot analysis (2 males, 1 female). Preliminary culture experiments revealed that microinjection did not impair in vitro development of ovine zygotes. These data indicate that microinjection of FVIII gene constructs into ovine zygotes is compatible with high embryonic survival rates, but the low percentage of transgenic offspring needs further study.

To interpret sensor signals like images, image sequences, or continuous speech the representation and use of task-specific knowledge is necessary. The paper presents a framework for the representation of declarative and procedural knowledge using a suitable definition of a semantic network. Based on that formalism a problem-independent control algorithm for the interpretation of sensor signals is presented. It provides both data-driven and model-driven control structures which can easily be combined to perform any mixed strategy. An explanation facility is available which makes the development of complex knowledge bases easier and increases the acceptance of such a knowledge-based analysis system., Unerläßliche Voraussetzung für die Interpretation von Sensordaten wie Bilder, Bildfolgen oder kontinuierliche Sprache ist die Darstellung und Nutzung von aufgabenspezifischem Wissen. In diesem Artikel wird ein Formalismus vorgestellt, der deklaratives und prozedurales Wissen in einem semantischen Netzwerk repräsentiert. Auf diesem Formalismus aufbauend ist ein problemunabhängiger Kontrollalgorithmus für die Interpretation von Sensordaten definiert. Er besitzt sowohl datengetriebene als auch erwartungsgesteuerte Kontrollmechanismen, wodurch jede gemischte Strategie erzielt werden kann. Eine Erklärungskomponente unterstützt die Entwicklung und Wartung komplexer Wissensbasen und erhöht die Akzeptanz eines solchen wissensbasierten Analysesystems.

A total of 917 cumulus-oocyte complexes were used in this experiment. In Group VI the cumulus-oocyte complexes were matured and fertilized in vitro. After an initial 48 hours culture in microdrops they were transferred to a granulosa cell monolayer. Development to morulae/blasto-cysts was evaluated 8 days after in-vitro fertilization. The treatment for Group I consisted of 50 ng/ml insulin-like growth factor I (IGF-I) added to in-vitro maturation media. Group II received IGF-I to in-vitro culture on the granulosa cell monolayer. In Group III of IGF-I was added to in-vitro maturation media and in-vitro culture on the granulosa cell monolayer. Group IV received IGF-I to in-vitro culture and was cultured the whole time without a granulosa cell monolayer. Group V received IGF-I to in-vitro maturation and were cultured in-vitro without a granulosa cell monolayer. After supplementation of in-vitro maturation media with IGF-I, 52, 49 and 51% of the cumulus-oocyte complexes in Groups I, III and V did not show cumulus expansion, whereas in Group VI (control) only 1% of the cumulus-oocyte complexes did not expand (P

In this paper the application of the finite element method is presented to compute the optical flow field. Using this approach homogeneous areas are roughly triangulated. And vice versa, a fine partition at motion boundaries ensures a flow field with high resolution. This drastically reduces the computational effort without loss of accuracy. We will describe the adaptation of the image partition to the variance of the gray level gradients as well as the basic steps for solving the variational problem. Experimental results are presented using synthetic and natural image sequences. Finally, the approach is compared to results computed by the finite difference method (grid method), used up to now.

We present a flexible environment for the generation of word hypotheses in continuous speech. After describing the interface to the other modules of our speech understanding system a verification algorithm and a word spotting technique both based on HMM will be discussed. The generation of reference models for the matching procedures is done automatically using the standard pronunciation of a word and a set of phonological rules about intra word assimilation. These alternative pronunciations are represented by graphs with labeled edges. Some preliminary results for the matching procedures are also given.

In this paper the performance of a nearly optimal system for character recognition is compared to human performance on the same data set. The recognition system uses a linear feature extraction method which is superior to discrete Karhunen-Loeve expansion. The experiments consider binary and multiple classification of handprinted characters, binary classification of similar characters of one font corrupted by additive white normal noise, and multiple classification of truncated handprinted characters. It turns out that the human visual system is superior in recognizing handprinted characters and inferior in the case of single font characters with additive noise.

Discrete Karhunen-Loeve-transform and a newly developed nonlinear mapping are used to obtain a structure-preserving mapping of high-dimensional data to a plane. It seems that the potential power of nonlinear mappings is needed only for artificial data.

The developmental capacity of frozen/thawed bisected embryos (n = 33) derived from day-7 bovine embryos was investigated and compared to ordinary embryos after freezing and thawing (n = 28) and to freshly bisected embryos (n = 19). The freezing and thawing protocol was identical for ordinary and demi-embryos. The percentage of intact embryos classified as excellent, good, or poor after thawing was 92.9 and 96.3% for ordinary and demi-embryos, respectively. Pregnancy rates of 53.8 ( 8 15 ), 46.2 ( 6 13 ), and 47.5% ( 9 19 ) were obtained when frozen/thawed ordinary embryos and frozen/thawed demi-embryos classified as excellent or good and sealed with an additional zona pellucida from hatched pig blastocysts or freshly bisected embryos were transferred. One pair of identical twins resulted from the transfer of frozen/thawed demi-embryos sealed with an additional zona pellucida. Transfer of four frozen/thawed demi-embryos without an additional zona pellucida led to one pregnancy. In contrast, demi-embryos derived from frozen/thawed ordinary embryos (n = 8) as well as frozen/thawed demi-embryos classified as poor (n = 6) did not result in any pregnancies although two halves were transferred per recipient. It is concluded that sealing the punctured zona pellucida improves the developmental capacity of frozen/thawed demi-embryos derived from day-7 bovine embryos, and freezing demi-embryos is more efficient compared to the splitting of frozen/thawed ordinary embryos.