Creating avatars using neural networks

What needs to be done We create watercolor illustrations based on client photographs. The subjects vary: portraits of people, animals, couples, families, people with animals, etc. We have a reference style, and your task is to consistently generate illustrations exactly in this style, without deviations, from any new photo regardless of the subject. Key requirements for the result: — Classic realistic watercolor: dense saturated pigment, multi-layered washes, visible brush strokes, texture of watercolor paper — Pure white background — paint only on the object, no stains or underlays behind — Portrait composition to the waist, with the figure smoothly dissolving into the white paper with an even gradient (no splashes, drops, or torn edges) — Object centered, with margins on all sides, nothing cropped by the frame — Complete resemblance to the photo — faces of people, appearance of animals, details of clothing (this is critical: clients order portraits of themselves, their loved ones, and their pets) — Vertical format 1792 × 2368 — Separate mode on request: full height (with legs/paws) Candidate requirements — Confident work with modern image models: Nano Banana 2 (Gemini), GPT Image 2, Midjourney (sref / Omni Reference), FLUX Kontext or analogs — Experience in reproducing a given style from a reference — show examples of “was photo — became in style X” — Understanding of prompt engineering for image models: control of composition, background, cropping — Ability to work with multi-image (style reference + client photo in one request) — Portfolio is mandatory Will be a plus — Experience working with commissioned portraits (facial resemblance) — Basic retouching skills (Photoshop / Procreate) for final refinement — Experience in building a stable pipeline: one prompt/workflow — same result on the flow, on different subjects Test task (paid, by agreement) We will send: 1 style reference + 2–3 photographs with different subjects (for example, a portrait of a person, an animal, a couple). You return: finished illustrations in the reference style for each photo. We evaluate according to the checklist: Style match with the reference (technique, saturation, character of strokes) — the same across all subjects Pure white background without foreign stains Smooth gradient dissolution at the bottom — without splashes and sharp cuts Composition: centered, nothing touches the edges of the frame Resemblance to the photo (faces, appearance of animals, details) Conditions — Payment for results (discussed: per illustration / package / monthly volume) — Volume: from N illustrations per week (indicate your actual pace) — Communication: Telegram Response: send portfolio + 2–3 examples of work “photo — stylization” + your pace and price per unit.

I'm looking for a new candidate for the project, as I couldn't find one last time. iPad + Procreate (strictly). Important content conditions: You are over 18 years old. You have a normal, non-hypocritical attitude towards drawing women and plus-size furry characters. This is our target theme, and you will be working with it constantly — please keep this in mind when responding. Who I am looking for (important criteria) I need not a "tracer," but a thoughtful artist who understands the logic of drawing. THIS IS REALLY IMPORTANT: Lines must connect, lines have thickness, thin in the middle and thick on the sides, with line thickening in places where it logically would be darker. Understands how and why the line goes this way and not otherwise. Thanks to the sequence of drawing lines, the logic of the drawing is established. Feels volume and structure: for example, that a leg is tucked into a boot (not the other way around), that the chest is tucked into a corset and therefore looks protruding, that the tail grows from the correct anatomical place, not from anywhere. Notices mistakes in proportions by themselves: too long arm, crooked face, missing fingers, etc. Can draw hands. Leads a smooth, lively line with correct thickening and thinning in the right places. Confidently works with color and references. Understands color combinations. Can replicate my line art and coloring style. What the task involves I will send a ready Procreate file with already set layers and groups. Your work consists of two stages: Line (line drawing) You will make a neat line art over the sent template. All lines must be complete, as coloring will be done through "Reference." You will show it to me for approval. Coloring (after line approval) The technique is simple, by layers for each object: Base color layer (Color) Clipping layers on top: Blushing (blush/shades) Add (adding light) Multiply (darkening/shadows) Speed: I spend 40–60 minutes on 1 drawing. I expect a comparable speed from you after mastering the technique. Payment: $10 for 1 drawing. Payment method: PayPal. How to respond and test task (please read carefully) No automatic responses and "I've been working in Photoshop for 250 years, I can take a picture of my mom" — save all that for the Nobel Prize. What needs to be done: Just show how you do the line. Send a line drawing — an outlined picture (cow) in your execution. Test task parameters: Brush: attached (.brushset), use it, brush size 4-5. Sheet size: 2500 x 3700 pixels. Examples of similar works are attached below. What I am looking at: If you can make a good, meaningful line — with coloring, I am sure you will manage without problems, I will help. If your line looks similar to mine (put your work next to mine and really look, does it look similar?), we can do a paid test work. Main mistakes of past candidates, please try to consider: I am posting this task again as a person was not found. Here’s what to pay attention to: Need two hands, not three. Why do I need a cow with three hands? I think it will be clear that this is AI. Objects are simply outlined without logic. If this is a mass of hair, it has large curls (inside which there are already small lines), imagine it as 3D and layer the curls on the head, not just outline the line because it goes here. I attached examples of what my line looks like, when you draw yours, put it next to mine and think, does it look like one person? (and I understand that someone will say it will still differ, I have worked for 7 years in cartoon portraits, and I know that you can adapt to a style with the right skills and knowledge). Please think of the objects as if you were drawing a character from scratch yourself. For example, the cow on the left has a cuff, there are under-folds and an under-flower, you need to draw it logically, not just outline it incorrectly. The same goes for the top, there are damaged folds, you draw regular stockings (which end with a line at the top), and then draw lace on top. Lines are even, not chaotic, look closely at my examples. Life hack, lower the brightness of my drawing and try to outline the lines to understand how to lead them correctly. If you have questions, write, I will help. I need a person who understands what looks wrong in the drawing and can correct it at their discretion. https://drive.google.com/file/d/1scOAcBs3GjRrhQgQBJlJY3RSmRqKEwrP/view?usp=sharing Here is a video example of how I lead the line.

Looking for a person who can create video creatives for advertising on TikTok Ads that sell products. I provide all the raw video materials. Your task is to edit them into an advertising creative that will perform well in the TikTok feed.What needs to be done Edit short videos (15–40 seconds). Create dynamic edits with a strong hook in the first 3 seconds. Add subtitles, text inserts, animations. Select trending music and sound effects. Make the videos as similar as possible to native TikTok content. Use different scripts and structures for testing.

Welcome! The company Adaptis needs to update graphic materials for A+ content on our website and on Amazon. We already have a fully prepared design for A+ content (cards, banners, infographics), but they depict a different shape of our product. We have changed the product design and now have its new 3D model. The main task is to create realistic renders of the new product shape and carefully "fit" them into the existing banners instead of the old shape. And immediately your cases

Developer for creating AI bots - multi-agent system and automations in AI application. What is needed: proficiency in n8n; experience integrating OpenAI and other AI models via API; working with webhooks, REST API, and Supabase; ability to create reliable AI workflows and automations. Create multi-agent systems from scratch or with a framework AI Agents, MCP, RAG