MIHOLA, D. Chytrá kamera pro sledování zájmových objektů [online]. Brno: Vysoké učení technické v Brně. Fakulta informačních technologií. 2025.
This is an excellent thesis, which goes beyond the scope of ordinary thesis both in terms of technical and experimental work.
| Kritérium | Známka | Body | Slovní hodnocení |
|---|---|---|---|
| Informace k zadání | The diploma thesis of the student David Mihola was very demanding both from the professional point of view, the complexity of design and implementation, and the extent of experimental work that was associated with testing, processing and evaluation of the results. I am therefore very satisfied with the results achieved. The student independently decided to extend the original assignment, which was focused only on the detection of objects of interest, also to include their recognition. To do this, he proposed a sequential processing of image information using machine learning algorithms that were optimized for low power allowing battery operation while maintaining the necessary system throughput. | ||
| Práce s literaturou | David actively and independently searched for the necessary literary sources, which he processed into a quality research from which he drew for his further work. | ||
| Aktivita během řešení, konzultace, komunikace | David was very active in solving the thesis, he initiated frequent consultation meetings with the supervisor, for which he always came perfectly prepared. | ||
| Aktivita při dokončování | The thesis was completed and properly tested well in advance, even with a number of improvements developed by the graduate beyond the scope of the assignment. | ||
| Publikační činnost, ocenění | The thesis has not yet been published as it is expected to be continued and published as part of the PhD studies. |
This is clearly a way above the average work. Its strongest point is clearly the implementation part where two different versions of embedded video capture and object detection were implemented and while the implementation was, given the nature of the devices, not easy, it was successful and works very well. The text part, even while some minor objections were reaise, is exceptional as well and so the only suitable evaluation of the thesis is "outstanding".
| Kritérium | Známka | Body | Slovní hodnocení |
|---|---|---|---|
| Náročnost zadání | To my opinion, the assignment of the thesis was quite demanding. THe student needed to study and obtain product information far beyond the standard duties of MSc level students at FIT. Moreover, the assignment required the student to perform extended analysis and experimental work. | ||
| Rozsah splnění požadavků zadání | I believe that the assignment of the thesis has not only been fulfilled but that it has been fulfilled with significant extension and creativity (e.g. subdivision of the machine learning based runtime between the camera and embedded systems or implementaion of in and experiments with two different plaforms). | ||
| Rozsah technické zprávy | The text part of the thesis is quite extensive and even without bibliography and apeendix it doncatins over 70 pages; the total size is 101 pages. | ||
| Prezentační úroveň technické zprávy | 90 | The presenatation level of the thesis is outstanding, However, minor objections can be raised regarding the size of the images that are in general small and hard to read (e.g. all of the Appendix A or Figure 6.1). Also, it would have been nice to summarize the results not only in numerical tables but to include some graphs (specifically in Chapter 6). It would have been nice to illustrate also the user interface (web interface) of the system which is quite nice but not shown in the text of the thesis. On more objection can also be raised regarding the fact that state of the art description is apparently somewhat "mixed" with the own contribution of the student whish makes in the principle difficult to evaluate the own work (e.g. Chapter 3, Section 3.5 contains most probably results of own experimental work). I believe that better practice is to separate state of the art reprt from the description of own work for clarity. | |
| Formální úprava technické zprávy | 90 | On the formal side, that thesis is also quite excellent. However, some minor objections can be raised also here. For example, the text of the thesis in many places contains "first person plural" expressions while the thesis must be individual work of the student. Also, some typography issues can be seen, such as e.g. the table of contents being in some parts 3rd level numbering while only 2nd level numbering occurs in other places. I believe a better practice is to have the numbering "regular". | |
| Práce s literaturou | 80 | The thesis contains a large number of references (103 in total) and clearly the work with the literature nas been on a very good level. However, the references themselves are often incomplete (making them impossible to identify of the provided web links fail/stop to work in the future). | |
| Realizační výstup | 100 | The implementation part of the thesis is clearly very outstanding and the video capture, object detection (in a nearly wearable battery powered device), passing the information to the server and presentation of the results on the web pages is very impressive. Also the formal side of the implementation, such as software exploitation and licensing conditions complies to the rules. | |
| Využitelnost výsledků | I truly believe that the results of the work are novel and very useful, useable in practice and in fact, nearly ready to be implemented in a commercial device and deployed. |
eVSKP id 165000