For a project I would strongly recommend a subject matter that you are curious or passionate about, as strong motivation increases the likelihood of completing the project.
Also consider your interests outside of CSE. If you like building physical things with an embedded software component, you might decide to acquire an Arduino or Raspberry Pi to create and operate your own vacuum robot, LIDAR, pulse oximeter, or license plate reader. If you are more inclined or restricted to things that can be done with just a laptop computer, you could explore pseudo-random number generators, hash functions, sorting networks, or factorization. These are just a few ideas off the top of my head.
In high school I wondered what happened under the hood when I pressed the $\exp$ key on my calculator, and could I implement that myself on an Apple II? After seeing someone recite one thousand digits of $\pi$ on television, I wondered how one would go about computing those 1000 digits of $\pi$. These were my first two software projects. Research was difficult as there was no internet yet, and I was mostly limited to what little the local public library and my school's library had to offer. I learned a lot and it created useful background knowledge for when I was later involved in the design of floating-point units and math libraries in my professional career.
Well-known examples of ambitious student projects are (1) the blindingly fast Pascal compiler Anders Hejlsberg developed that later turned into the product Turbo Pascal. It cleverly rolled what are normally multiple stages in classical compiler construction into a single pass (2) Linus Torvalds' development of Linux, as he was curious what it would take to create his own Unix-like operating system for his new 386-based PC.
To ensure that other people's approaches are not cramping your creativity, I would recommend to first think hard about how to tackle the problem you posed yourself, implement that, and experiment with it. Failure to achieve any intermediate objective will provide a learning opportunity. The old adage applies: "If at first you don't succeed, try, try, try again".
After a while of applying your own ideas, start trawling the literature for relevant information. You will learn how to research independently by following both backward pointers (references cited by the publication you are reading) and forward pointers (publications that cite the publication you are reading). I would also encourage putting into practice some of the methods and algorithms found in the literature. The claims made are sometimes misleading, inaccurate, or downright false.
I am a retired software engineer who looked at plenty of resumes in his time. Many resumes from new college grads look very much alike, and these days it is common to see some projects listed. I would be looking for anything that indicates a deeper interest in (or excitement for) the subject matter, creativity, and persistence in pursuing a working software product. Plus points for a comprehensive, well-designed test framework, which is harder to achieve and requires more creativity than some may assume.
