| 1. |
|
|
| 2. |
- Gustafsson, Ny Björn, 1975-
(författare)
-
Casting Identities in Central Seclusion : Aspects of non-ferrous metalworking and society on Gotland in the Early Medieval Period
- 2013
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- The aim of this thesis has been to investigate and interpret late Iron Ageand Early Medieval traces of non-ferrous metalworking on the islandGotland, Sweden. Gotland was not, based on the archaeological record, anintegrated part of the common Scandinavian culture. Instead a local,endemic cultural expression had developed; a seclusion which lasted forcenturies despite the islands central position in the Baltic Sea. In thepast, key elements for the understanding of local settlement- and burialpractices as well as the local material culture were mainly recovered andreported by local farmers. A specific category of such finds – so-called‘bronze slag’ is discussed and partly reinterpreted in the first study ofthis thesis. Two further studies treat different aspects of metalworkingand metalworkers – one discusses common archaeological notions ofScandinavian workshops, production sites and metalworkers from a criticalperspective while the other mainly focuses on the Gotlandic finds frommetal-detector surveys carried out over the last 35 years. Based on whereand to which extent, both from a quantitative and a qualitative point ofview, these finds occur a hierarchical classification into four sub groupsis presented – ordinary farm sites with traces of non-ferrous metalworking,workshop sites, potential workshop sites and last, extrovert harboursettlements. A fourth study presents an attempt to evaluate the usefulnessof magnetometry in delimiting extant traces of high-temperature crafts,such as metalworking. The last study of the thesis presents an attempt touse trace elements analysis of skeletal lead in human bone to identifypotential non-ferrous metalworkers.As the wearing of endemic Gotlandic jewellery appears to have been centralin the manifestation of the local identity it is argued that themetalworking artisans played a crucial role in defining how this identitywas signalled and displayed via the jewellery and dress-related metalobjects. It is further suggested that these artisans might have played animportant role in upholding the local economy before the advent of localminting.
|
|
| 3. |
- Gustin, Ingrid, et al.
(författare)
-
Elites, networks and the Finnish connection in Birka
- 2016
-
Ingår i: New Aspects on Viking Age Urbanism c. AD 750-1100 : Proceedings of the International Symposium at the Swedish History Museum, April 17-20th 2013 - Proceedings of the International Symposium at the Swedish History Museum, April 17-20th 2013. - 1102-1195. - 9789189338203 ; 12, s. 49-61
-
Bokkapitel (refereegranskat)abstract
- In research on Birka strong focus has since long been directed towards long-distance contacts with Western as well with and Northeastern Europe (todays Russia, Belarus and Ukraine). The great attention paid to west- and east European contacts has likely contributed to that contacts with Finland have fallen out of the discussions. However as seen from studies of the pottery in Birka the Baltic Finnish ware constitute the second largest group when it comes to imports to the town. In the graves on Björkö there are finds of this type of ware as well as of dress details deriving from or showing heavy influence from the Finnish mainland. Some of these finds derive from the most lavishly furnished chamber graves in Birka, indicating that the elite in the town were part of a network that was directly or indirectly connected to groups on the Finnish mainland. Besides Baltic Finnish pottery or dress details from Finland these graves also contained objects such as weights, scales and Islamic coins – items indicating an affiliation to groups involved in the sale and exchange of goods. Finds of standardized weights in south western Finland is a clear indication that this region was part of a supra regional network in the Viking Age just like Birka. Objects from Finland found in Birka on one hand and objects from Scandinavian and central Sweden found in Finland on the other confirm that there has been a close connection between the Mälaren region and south western Finland. Probably products such as furs distributed by the elite in Vakka Suomi attracted groups in eastern Sweden and Birka. However the elite in Birka might also have had other reasons for maintaining contacts with groups in south western Finland. People in south western Finland were linked to routes and networks spanning much further east and by extension also to the riches of the Caliphate.
|
|
| 4. |
- Hedenstierna-Jonson, Charlotte, 1971-
(författare)
-
The Birka Warrior : the material culture of a martial society
- 2006
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- This is a study of martial material culture in the context of the Viking Age warrior of Birka, Sweden. The aim is to establish the role, function and affiliation of the Birka warrior and thereby place Birka on the power-political map of the 10th century. The study is based on the excavations of the fortified structures, particularly the Garrison, at the trading post of Birka as well as the extensive remains of material culture deriving from these investigations. A starting hypothesis is that an analysis of material culture constitutes a way of mapping social structures and that style and iconography reflect cultural groups, contacts and loyalties.Based on the case studies of six papers, the synthesis deals with questions of the work and world view of the warriors, as too their relation to their contemporary counterparts in eastern and western Europe. Questions are raised concerning the value and function of symbols in a martial context where material culture reflects rank, status and office. In defining the Birka warrior’s particular stylistic expression, a tool is created and used in the search for contacts and affiliations reflected through the distribution patterns. The results show close contacts with the eastern trading posts located on the rivers Volga and Dnjepr in Ancient Russia.It is stated that these Rus’ trading posts, essentially inhabited by Northmen, shared a common cultural expression that was maintained throughout a vast area by exceptionally close contacts. It is suggested that a particular stylistic expression developed in these Rus’ trading places containing elements of mainly Scandinavian, Steppe nomadic and Byzantine origin.In conclusion, the results of this thesis show that the warriors from Birka’s Garrison had a share in the martial development of contemporary Europe but with their own particular traits. Close relations with the eastern trade route and contact with the powerful Byzantine Empire were enjoyed. As a pointer for future research, it is wondered what organisational form the close-knit structure of the Rus’ trading posts actually took, keeping the subsequent guilds of medieval Europe in mind. The fall of the Garrison, as of Birka, corresponds with the establishment of Christianity in the region. Such changes were not limited to Central Sweden but part of a greater process where a new political structure was developing, better anchored in local concerns.
|
|
| 5. |
|
|
| 6. |
- Holmquist, Karl, 1992-, et al.
(författare)
-
Class-Incremental Learning for Semantic Segmentation - A study
- 2021
-
Ingår i: 2021 Swedish Artificial Intelligence Society Workshop (SAIS). - : IEEE. - 9781665442367 - 9781665442374 ; , s. 25-28
-
Konferensbidrag (refereegranskat)abstract
- One of the main challenges of applying deep learning for robotics is the difficulty of efficiently adapting to new tasks while still maintaining the same performance on previous tasks. The problem of incrementally learning new tasks commonly struggles with catastrophic forgetting in which the previous knowledge is lost.Class-incremental learning for semantic segmentation, addresses this problem in which we want to learn new semantic classes without having access to labeled data for previously learned classes. This is a problem in industry, where few pre-trained models and open datasets matches exactly the requisites. In these cases it is both expensive and labour intensive to collect an entirely new fully-labeled dataset. Instead, collecting a smaller dataset and only labeling the new classes is much more efficient in terms of data collection.In this paper we present the class-incremental learning problem for semantic segmentation, we discuss related work in terms of the more thoroughly studied classification task and experimentally validate the current state-of-the-art for semantic segmentation. This lays the foundation as we discuss some of the problems that still needs to be investigated and improved upon in order to reach a new state-of-the-art for class-incremental semantic segmentation.
|
|
| 7. |
- Holmquist, Karl, 1992-
(författare)
-
Data-Driven Robot Perception in the Wild
- 2023
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- As technology continues to advance, the interest in the relief of humans from tedious or dangerous tasks through automation increases. Some of the tasks that have received increasing attention are autonomous driving, disaster relief, and forestry inspection. Developing and deploying an autonomous robotic system to this type of unconstrained environments —in a safe way— is highly challenging. The system requires precise control and high-level decision making. Both of which require a robust and reliable perception system to understand the surroundings correctly. The main purpose of perception is to extract meaningful information from the environment, be it in the form of 3D maps, dense classification of the type of object and surfaces, or high-level information about the position and direction of moving objects. Depending on the limitations and application of the system, various types of sensors can be used: lidars, to collect sparse depth information; cameras, to collect dense information for different parts of the visual spectra, of-ten the red-green-blue (RGB) bands; Inertial Measurements Units (IMUs), to estimate the ego motion; microphones, to interact and respond to humans; GPS receivers, to get global position information; just to mention a few. This thesis investigates some of the necessities to approach the requirements of this type of system. Specifically, focusing on data-driven approaches, that is, machine learning, which has been shown time and again to be the main competitor for high-performance perception tasks in recent years. Although precision requirements might be high in industrial production plants, the environment is relatively controlled and the task is fixed. Instead, this thesis is studying some of the aspects necessary for complex, unconstrained environments, primarily outdoors and potentially near humans or other systems. The term in the wild refers exactly to the unconstrained nature of these environments, where the system can easily encounter something previously unseen and where the system might interact with unknowing humans. Some examples of environments are: city traffic, disaster relief scenarios, and dense forests. This thesis will mainly focus on the following three key aspects necessary to handle the types of tasks and situations that could occur in the wild: 1) generalizing to a new environment, 2) adapting to new tasks and requirements, and 3) modeling uncertainty in the perception system. First, a robotic system should be able to generalize to new environments and still function reliably. Papers B and G address this by using an intermediate representation to allow the system to handle much more diverse types of environment than otherwise possible. Paper B also investigates how robust the proposed autonomous driving system was to incorrect predictions, which is one of the likely results of changing the environment. Second, a robot should be sufficiently adaptive to allow it to learn new tasks without forgetting the previous ones. Paper E proposed a way to allow incrementally adding new semantic classes to a trained model without access to the previous training data. The approach is based on utilizing the uncertainty in the predictions to model the unknown classes, marked as background. Finally, the perception system will always be partially flawed, either because of the lack of modeling capabilities or because of ambiguities in the sensor data. To properly take this into account, it is fundamental that the system has the ability to estimate the certainty in the predictions. Paper F proposed a method for predicting the uncertainty in the model predictions when interpolating sparse data. Paper G addresses the ambiguities that exist when estimating the 3D pose of a human from a single camera image.
|
|
| 8. |
- Holmquist, Karl, 1992-, et al.
(författare)
-
Evidential Deep Learning for Class-Incremental Semantic Segmentation
- 2023
-
Ingår i: Image Analysis. SCIA 2023.. - : Springer. - 9783031314377 - 9783031314384 ; , s. 32-48
-
Konferensbidrag (refereegranskat)abstract
- Class-Incremental Learning is a challenging problem in machine learning that aims to extend previously trained neural networks with new classes. This is especially useful if the system is able to classify new objects despite the original training data being unavailable. Although the semantic segmentation problem has received less attention than classification, it poses distinct problems and challenges, since previous and future target classes can be unlabeled in the images of a single increment. In this case, the background, past and future classes are correlated and there exists a background-shift.In this paper, we address the problem of how to model unlabeled classes while avoiding spurious feature clustering of future uncorrelated classes. We propose to use Evidential Deep Learning to model the evidence of the classes as a Dirichlet distribution. Our method factorizes the problem into a separate foreground class probability, calculated by the expected value of the Dirichlet distribution, and an unknown class (background) probability corresponding to the uncertainty of the estimate. In our novel formulation, the background probability is implicitly modeled, avoiding the feature space clustering that comes from forcing the model to output a high background score for pixels that are not labeled as objects. Experiments on the incremental Pascal VOC and ADE20k benchmarks show that our method is superior to the state of the art, especially when repeatedly learning new classes with increasing number of increments.
|
|
| 9. |
- Holmquist, Lena
(författare)
-
700-tal - älgmannen och ynglingen
- 2020
-
Ingår i: Begravd på Birka. - : Birka Vikingastaden. - 9789151944821
-
Bokkapitel (populärvet., debatt m.m.)
|
|
| 10. |
|
|
